You are on page 1of 70

1

Particulate filled polypropylene:


structure and properties
B. Pukanszky

1.1 INTRODUCTION
Particulate-filled polypropylene (PP) has been used in large quantities in
numerous fields of applications for many years. Figure 1.1 shows not only
the most important applications, but also the obviously high growth rate of

1986 1990

45

25

50


Automotive Furniture Electrical,
Business
Domestic OPP, fibers
tapes, etc

Figure 1.1 The most important fields of application for particulate filled PP (values
in 1000 tons).

Polypropylene: Structure, blends and composites. Edited by 1. Karger-Kocsis.


Published in 1995 by Chapman & Hall, London. ISBN 0 412 58430 1
2 Particulate jilled polypropylene: structure and properties
these materials [1]. The success of particulate-filled PP lies in its extremely
advantageous price/volume/performance relations [2-5], with the result
that PP composites successfully penetrate fields traditionally occupied by
other materials such as ABS [6,7]. Considerable efforts have been made to
extend their application to fields where engineering thermoplastics have
been used exclusively up to now [8,9].
Although numerous different fillers and reinforcements are added to PP,
three playa dominant role and will continue to do so in the future. While
CaC0 3 takes an overwhelming proportion of the filler market in plastics, in
PP it occupies second place behind talc (Table 1.1) [1,2]. The third filler, or
more exactly reinforcement, is glass fiber, which is used in lesser quantities.
Intensive research and development is dedicated to glass fiber reinforced PP,
to improve its performance and to extend its application [8-11]. Despite
these efforts, potentials and predictions, both the actual quantity and the
growth rate of glass fiber consumption in PP are inferior to those of the
principal particulate fillers (Figure 1.2) [1].
Originally fillers were introduced to extend the polymer and to decrease
the price of the compound. It must be emphasized, however, that filled
polymers possess numerous advantages over their non-modified counter-
parts. Among others, fillers increase stiffness and heat deflection tempera-
ture, decrease shrinkage and improve the appearance of the composites [1,
4-6, 12]. Through their decreased specific heat and increased heat conduc-
tivity, productivity can be increased in most processing technologies [4, 5].
Fillers are very often introduced into the polymer because of some new
functional properties, like flame retardance or conductivity, not possessed

Table 1.1 West European market for fillers and reinforcements for plastics and
PP (1000 tons)

Plastics [2] PP [1]

Filler type 1986 1987 1990" 1993"

Calcium carbonate 845 34.0 48.0 72.0


Talc 85 43.0 55.3 74.0
Kaolin 36
Aluminum hydroxide 36
Asbestos 12
Glass fibers 7.0 8.5 11.0
Silica 6 Small Small Small
Wollastonite 2 0.2 0.4 0.5
Mica 0.5-1 1.9 2.4 2.8
Glass beads 0.3 0.4 0.5
Wood fibers 0.1 0.2 0.2

"estimated
Introduction 3
100.-----r---~----~----._--~

I 8O--------i-----t----~~---i
~ ----T---~---
60

"s.
E
iil 40 .......... ...•.........................;......................... j .........................
:s ! CaC0 !
i
3 i

I [ i
i< 20 ------t--;;;;;;Ifi':--l--:r=--
i
O+---~~---+----~----+_--~
!
1985 1987 1989 1991 1993 1995
Year
Figure 1.2 Amount and growth of filler consumption in Western Europe for the
fillers and reinforcements used in the largest quantities in PP.

by the matrix polymer at all, thus resulting in principally new composites


[13,14]. The ever-increasing technical and aesthetic requirements, as well as
soaring material and compounding costs, lead to the necessity of utilizing
all possible advantages of fillers.
To meet these requirements - utilizing all the functional advantages of
fillers or using them only as extenders - it is necessary to understand the
relevant properties of particulate fillers and their effect on the structure and
properties of the composites [1,5,15]. This chapter, therefore, focuses on
the most important questions which must be considered during the devel-
opment of a new PP composite. These issues will be discussed in the
following order:
1. Effect of fillers on the crystalline structure of PP; factors influencing the
properties of particulate filled PP.
2. Stress distribution and micromechanical deformations in PP composites.
3. Effect of interfacial interactions on properties.
4. Modification of interactions, surface treatment.
5. Homogenization, processing and structure of particulate filled PP.
6. Estimation and prediction of properties.
7. Multicomponent PP systems, hybrids, simultaneous elastomer and filler
modification.
The relation of particulate fillers and glass fiber reinforcements must
also be mentioned briefly. The transition from fillers to reinforcements is
4 Particulate filled polypropylene: structure and properties
continuous; it is quite difficult to draw the line. There are also numerous
similarities in the factors determining the structure and properties of the
composites. The micromechanical deformation processes are basically the
same, as are the role of particle geometry (aspect ratio), particle orientation
(e.g. talc, mica, cellulose, etc.) or surface treatment. Although in this chapter
attention will be focused on particulate fillers, because of these similarities
occasional reference will also be made to the fiber reinforced systems.

1.2 EFFECT OF FILLERS AND REINFORCEMENTS ON


PP MATRIX STRUCTURE
PP is a semi-crystalline polymer. Its properties are determined by its
crystalline structure, and the relative amount of amorphous and crystalline
phases, crystal modification, size and perfection of crystallites, dimensions of
spherulites and the number of tie molecules all influence the performance of
PP products [16]. Nevertheless, direct correlations between a particular
morphological characteristic (e.g. crystallinity, spherulite size, etc.) and
macroscopic properties are difficult to determine, since with changing
crystallization conditions or during annealing most of these characteristics
change simultaneously. It seems to be certain that with increasing crystal-
linity and spherulite size, the modulus of PP increases, and its strength and
especially its deformability decrease. The correlation was proved by Greco
and Coppola [17] (Figure 1.3), who have shown the importance of tie
45

40 ..... .........
I
co 35 I
a..
~ /
.s:: 30 ~

::,
c;,
c: ~
~
ti 25 ~
..!E
'Cii
c:
CIl
20
V
I-

15

10
o 2.5 5 7.5 10 12.5 15
Number of tie molecules x 10- 18 (C-Ccm-2)
Figure 1.3 Dependence of tensile strength on the number of tie molecules In
polypropylene [17].
Effect of fillers and reinforcements on PP matrix structure 5
molecules in the mechanical behavior of PP. The correlation was also
quantitatively verified by Wright et al. [18].
The structure of injection molded PP products is especially interesting
and important. The high shear and the thermal conditions produce a
multi-layered skin-core structure, in which the type, size and orientation of
both the crystalline and amorphous phases change continuously from the
skin towards the core [19-22]. Close correlation has been found between
structure and properties [23,24J; the highly sheared layer containing a
significant number of f3 spherulites is the weakest point, where crack
initiation starts, resulting in poor impact strength [25,26J and decreased
fatigue resistance [24, 27J.
Introduction of a second component - filler or reinforcement - into the
polymer can change its crystalline structure, which, as a consequence, results
in property differences [28, 29]. The most important effect of particulate
fillers is their ability to act as nucleating agents. The very strong nucleating
effect of talc has been demonstrated repeatedly [30,31]. Strong nucleation
can lead to a change of crystal modification. Introduction of talc into the f3
modification of PP resulted in a complete change of crystalline structure: the
higher crystallization temperature of the IX modification prevented the
formation of the f3 form [32-34].
Although the nucleating effect of talc has been unambiguously proven,
the influence of other fillers and reinforcements is not so clear. Many fillers
have shown weak nucleation effect [35-37J, while some of them leave the
crystalline structure of PP untouched [38, 39J. However, the definition and
classification of nucleating agents is not clear either. Varga [32J classified
talc as an active filler, and CaC0 3 , carbon black and dolomite as inactive
ones. Fujiyama and Wakino [31J, on the other hand, used CaC0 3 as a
nucleating agent in their studies. Changing the particle size of CaC0 3 ,
however, leads first to the appearance of a second crystallization peak, then
to the shift of the complete crystallization exotherm to higher temperatures,
showing a very strong nucleation effect of this filler (Figure 1.4).
The reasons for these differences in the nucleating efficiency of fillers
and reinforcements are not completely clear. Experiments conducted by
Ribnikat [36J indicated that changes in surface chemistry influence the
behavior significantly; treatment of the surface of a calcite crystal with
different chemicals led to changes in the nucleation effect. Energetics might
be another reason for these differences. Talc has a lower surface free energy
(140 m] jm 2) than the other fillers. On the other hand, surface treatment of a
filler with an organic compound results in a drastic decrease of surface
tension which is very often not accompanied by a change in nucleation
efficiency. Examples have been given in the literature where surface treat-
ment did not influence nucleation [40, 41J, while in other cases it decreased
or ceased completely on the coverage with some coupling agent [35, 37,42].
A closer scrutiny of literature information and experimental data indicates
6 Particulate filled polypropylene: structure and properties
16

14
b) f\

\
.. \
;:
o 6 f\ c)
::\
:::.~.. ..\
:;:
iii
Q)
::c 4
d) a)

l\
2

, \
C:l-~
o- ..-/1 ) '
150 130 110 90
Temperature (OC)
Figure 1.4 Influence of particle size and aggregation on the nucleation ability of
CaC0 3 fillers. Filler content: 20 vol%, Specific surface area (Ad, (a) no filler, (b)
2.2 m 2 jg, (c) 5.0 m 2 jg, (d) 16.5 m 2 jg.

that mostly physical and especially topological factors determine the nu-
cleation effect of fillers. In connection with the data of Figure 1.4 it has been
shown that the significant increase of nucleating effect in the case of CaC0 3
was connected to aggregation [35], which increases with decreasing particle
size. Surface treatment results in a decrease of aggregation, on one hand,
while thicker interlayers change the topological character, on the other
[35, 37, 42]. Both result in changes of the nucleating effect of fillers.
Further proof of the importance of topological effects is the transcrystall-
ization phenomenon of PP. On anisotropic filler particles or reinforcements
row nuclei can form and in such cases spherulites grow epitaxially to the
surface of the particle [36, 40, 41, 43, 44]. Transcrystallinity* is initiated by
the orientation of polymer molecules induced by shear or other topological
effects [45,46]. Properties of the transcrystalline layer differ from those of
the spherulitic form of PP; it possesses higher rigidity and lower defor-
mability, which lead to easier crack initiation and propagation [29,44].
Significant orientation of anisotropic filler particles may lead to a larger
amount of oriented transcrystalline polymer phase and to a drastic change
of composite properties.
Similarly, it is difficult to establish the effects exerted on composite pro-
perties by changes in the crystalline structure, even in the case of non-

* Also treated in Chapter 3 of Volume 1 and Chapter 10 of Volume 3.


Effect of fillers and reinforcements on PP matrix structure 7
oriented, spherulitic morphology. This effect is neglected in many cases.
Kendall [47] claims that changes in matrix structure do not influence
impact properties of particulate filled composites, and that the determining
factor is the strength of matrix-filler interaction in both amorphous and
crystalline polymers. In other cases close correlation was found between the
crystallinity or some other crystallization characteristic of PP and composite
properties [29,48-51]. Rutley and Darlington [50,51] found a more or less
linear correlation between both the crystallization temperature and the rate of
crystallization and the falling weight impact strength of particulate filled PP,
while Maiti and Mahapatro [49] observed an even better linear correlation
between the crystallinity and tensile characteristics of CaC0 3 filled PP
(Figure 1.5). Fujiyama and Wakino [31], on the other hand, related changes
in the properties of injection molded PP composites to the orientation of the
crystalline phase. Introduction of a filler into the PP matrix, however, changes
many factors simultaneously, which, moreover, have dissimilar dependence
on composition (Figure 1.6) and filler charac-teristics. These latter significant-
ly influence composite properties even in amorphous polymers; thus, separ-
ation of the influencing factors (crystal size, spherulite size, crystallinity,
interfacial interaction, effect of particle characteristics, etc.) is extremely
difficult. A similar effect of a filler on two or more properties of the composite
might lead to correlations like that presented in Figure 1.5.
For more information on the effects of fillers and reinforcements on the
PP matrix, see Chapters 3, 6 (Volume 1) and 4, 10 (Volume 3).

40.---------~--------.-------~

c;; 30
a..
::2:
'"
'"
!!'
1;;
"0
20
a;
'>'
~
'iii
c
<I>
I- 10

O+---------+---------+-------~
30 50 70 90
Heat of crystallization (Jig)
Figure 1.5 Effect of matrix crystallinity on the tensile yield stress of PPjCaC0 3
composites [49].
8 Particulate filled polypropylene: structure and properties
150 140

140 130
~-o

.~
...........?. . )

/
130 120

,:@
110 -;:;

,
I
,..~ ..... <I

110 . . . . . . .Af ~ ...

/
100

100 90

90 80
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.6 Changes in crystallization peak temperature and crystallinity as a
function of composition in PP/talc composites: (0) crystallization peak temperature
(Tc), (6) heat of crystallization (LiHc).

1.3 EFFECT OF FILLER CHARACTERISTICS ON COMPOSITE


PROPERTIES
The effect of filler characteristics on the performance and properties of PP
composites is at least as contradictory as their influence on crystalline
structure. Schlumpf and colleagues [4, 5, 52J have pointed out in several
publications that the chemical composition of fillers, which is usually
supplied by the producer as relevant information, is not sufficient for their
characterization. Further physical, mostly particle, characteristics are
needed to forecast their performance in a composite for any application [5].
The importance of obtaining general information about the effect of differ-
ent filler characteristics on composite properties becomes especially obvious
if we consider the large variety of materials used, or at least experimented
with, as fillers in PP composites.
Although the fillers used in the largest quantities in PP are talc and
CaC0 3 (see Table 1.1 and Figure 1.2) a great number of other materials are
also mentioned as potential fillers or reinforcements. Some of the most
frequently used materials are mica [29,37, 41J, short [22,43, 44J and long
glass fibers [53, 54J, glass beads [55, 56J, sepiolite [57, 58J, magnesium
hydroxide [59,60J, wood flour and cellulose [28,61]. Wollastonite [56J,
gypsum [62J, carbon black [63J, clay [29J, metal powders (aluminum, iron,
nickel) [48,64J, steel fibers [65J, silicium carbide [64J, phenolic micro-
Effect of filler characteristics on composite properties 9
spheres [38J and diverse flame retardants [39J are also mentioned. The
chemical variety of the fillers is obvious, and this often leads also to diffe-
rences in particle characteristics.
Chemical composition and, especially, purity of the filler have both a
direct and an indirect effect on its application possibilities and performance.
Traces of heavy metal contamination decrease the stability of PP [5,66].
Insufficient purity leads to discoloration of the product and limits the
application of the filler. High purity CaC0 3 has the advantage of white
color, while the grey shade of talc-filled composites excludes them from
some application fields. The chemical composition of the filler surface has
some influence on its nucleation effect [36]. Filler is surface treated in most
applications. It has been shown that acidity of the surface is crucial in silane
treatment of fillers [67J, and coupling agents must be chosen according to
the chemical characteristics of both the filler and the polymer matrix.
Mechanical properties of PP composites containing non-treated fillers are
determined mainly by their particle characteristics. One of the basic pieces of
information supplied by the manufacturer is the average particle size of the
filler. Although some contradictory data [68J can be found in the litera-
ture, it seems to be evident that particle size has a pronounced influence on
composite properties [5,69-71]. Strength, and sometimes modulus, in-
crease, deformability and impact strength decrease with decreasing particle
size.
Particle size in itself, however, is not sufficient to characterize any filler. As
was correctly pointed out by Schlumpf [5], knowledge of the particle
size distribution is equally important. Large particles - top cut - beside
changing abrasion and appearance of the product, usually have a strong
adverse effect on the deformation and failure characteristics of the com-
posites. The volume in which stress concentration is effective is said to
increase with particle size [71J and matrix-filler adhesion also depends on it
[55].
The other end of the particle size distribution, i.e. the number of small
particles, is equally important. The aggregation tendency of fillers increases
with decreasing particle size [72,73]. Extensive aggregation, however, leads
to insufficient homogeneity, rigidity and lower impact strength [74]. Ag-
gregated filler particles act as crack initiation sites in impact [29, 70, 74].
The particle size distribution of fillers is usually determined by sedimenta-
tion techniques. This, however, can also be misleading. In Figure 1.7 the
particle size distribution curves of two fillers are shown. Both fillers have
some tendency for aggregation - they have a fraction of small particles - and
the particle size distribution curves determined using sedimentation and
microscopy differ significantly. These differences also appear in the proper-
ties.
The specific surface area of fillers is closely related to their particle size
distribution. It also has, however, a direct impact on composite properties.
10 Particulate filled polypropylene: structure and properties
60 .---------~--------,_------_.

microscopy

40

>-
u
c:
Q)
:J
55 sedimentation
u:
20

o 5 10 15
Particle size (Ilm)
Figure 1.7 Particle size distributions of fiBers showing a tendency for agglomer-
ation. Dependence of distribution on the method of determination: (0) CaC0 3, (L'»
dolomite.

Adsorption of both small molecular weight additives and the polymer is


proportional to the size of the matrix-filler interface [5,15,52]. Adsorption
of additives may change stability, while matrix-filler interactions significant-
ly influences mechanical properties, particularly yield stress, tensile strength
and impact resistance [75,76].
The shape of the particles has pronounced significance. Reinforcement
increases with anisotropy of the particle; in fact, fillers and reinforcements
are very often differentiated by their degree of anisotropy (aspect ratio).
Plate-like fillers such as talc and mica reinforce PP more than spherical
fillers, and the influence of glass fiber is even stronger (Figure 1.8) [5,52,56].
In the case of anisotropic fillers it is difficult to obtain a clear picture about
the effect of particle characteristics on composite properties. Modulus seems
to increase with aspect ratio [29, 77], although Parrinello [78] found the
modulus of short glass fiber filled PP to be independent of both the length
and the diameter of the fibers. According to Riley et aT. [29], impact
resistance increases with decreasing particle size, large particles act as flaws,
while large aspect ratio results in increased stress concentration. Vu-Khanh
and Fisa [79], on the other hand, observed a decrease in the absorbed
impact energy with decreasing size in spite of some increase in the crack
initiation resistance. Tensile strength is said to decrease with particle size,
although Trotignon et aT. [80] did not observe any change in its value as a
function of this characteristic. In spite of these contradictions, or maybe
Effect of filler characteristics on composite properties 11
6~----------------------~------.

5 -j. ............................... ;................................... ; ................................ +v


-;0
a..
<.!)

II)
4
:l
:l
"tl
0
E
3
~
:l
X
Q)
iI
2

0.05 0.1 0.15 0.2


Volume fraction of filler ( - )
Figure 1.8 Effect of filler anisotropy on the flexural modulus of PP composites
[5]: (0) glass fiber, (L';) talc, (0) CaC0 3.

because of them, it is clear that beside particle size and particle size
distribution, aspect ratio and its distribution must also be taken into
consideration in the application of a new, anisotropic filler.
Hardness of the filler has a strong effect on the wear of the processing
equipment [5, 12]. Hardness, however, is not the only factor that determines
equipment wear; size and shape of particles, composition, viscosity and
speed of processing also play an important role [5].
Surface free energy of the fillers determines both matrix-filler and par-
ticle-particle interactions. The former has a pronounced effect on the
mechanical properties of the composite, the latter determines aggregation
[5,81]. Both interactions can be modified by surface treatment. Non-
reactive treatment leads to improved dispersion, but decreased matrix-filler
interaction [35,73,81], while chemical or physical coupling results in
improved strength (Figure 1.9) [82, 83]. Some fillers and reinforcements are
supplied in a surface-treated form. The amount and character of the
treatment must be known for successful application of the filler.
Thermal properties of fillers differ significantly from those of PP. This has
a beneficial effect on productivity and processing: decreased heat capacity
and increased heat conductivity decrease cooling time [5]. Changing ther-
mal properties of the composites results in a change of the crystalline
skin-core morphology and, thus, the properties of injection molded parts as
well. Large differences in the thermal properties of components, on the other
hand, lead to the development of thermal stresses [84,85], which also
influence the performance of the composite under external load.
12 Particulate filled polypropylene: structure and properties
40.-----,-----r-----.----.-----.

<0
Q.

~ 30

~
ti

i I
~ 20 ...-....................l.. . . . . . . . . . . . ~. . . .. .1..........................
~ I!
!
;
j
I I
!
!
10+-----~--~----~----1---~
o 0.1 0.2 0.3 0.4 0.5
Volume fraction of filler (-)
Figure 1.9 Effect of surface modification and coupling on the tensile yield stress of
PP/CaC0 3 composites: (0) non-treated, (0) stearic acid treated, (L\.) MA-PP modi-
fied.

Some fillers are added to PP to achieve some functional properties not


possessed by the polymer itself. The most important of these are flame
retardants and conductive fillers [13,14,39,59,60]. Particle characteristics
and physical properties of these materials exert the same influence on the
properties of the composites as those of the simple extenders. Complete
characterization of these fillers is absolutely essential to achieve the desired
goal, of preparing composite with the given functional property and accept-
able mechanical and aesthetic characteristics at the same time.
The filler content in PP is usually considerable both in the case of
extenders and functional fillers. Very often the latter even have a threshold
composition below which they are not effective [86J. Composite properties,
however, depend strongly on composition; the dependence is usually a
function of particle characteristics. With increasing filler content stress
distribution in the composite changes and with the association of filler
particles higher order structures can form. Performance of a filler in any
composite, therefore, must always be determined as a function of filler
content.

1.4 STRESS ANALYSIS, MICROMECHANICAL DEFORMATIONS


Introduction of fillers or reinforcements into the polymer matrix results in a
heterogeneous system. Under the effect of external load these hetero-
Stress analysis, micromechanical deformations 13
geneities induce stress concentration, the magnitude of which depends on
the geometry of the inclusions, on the relative properties of the com-
ponents and on interfacial adhesion [87,88]. Heterogeneous stress dis-
tribution and local stress maximums developing in the composite obvious-
ly influence its deformation and failure behavior as well as its overall
performance.
Although the importance of stress concentrations in PP composites has
been pointed out in numerous publications, their exact role is not com-
pletely clear. Nakagawa and Sano [89] claim that an increased number of
homogeneously distributed stress concentration sites created by well dis-
persed filler particles leads to an increase of impact strength. Their state-
ment is supported by the results of Michler and Tovmasjan [70J, who
found that homogeneously distributed small filler particles initiated craz-
ing, which led to increased energy absorption and improved impact resis-
tance. According to others, stress concentration causes a deterioration in
properties, the initiated micromechanical deformations result in decreased
strength and impact resistance [29,90]. Trantina [91], on the other hand,
found in model experiments that holes punched in a PP film did not act as
stress concentrators, and that stresses could be calculated on the basis of
effective matrix cross-section. Similarly contradictory is the relation of
particle dimensions to stress concentrations. Small particles seem to be
more advantageous, because in the case of large particles a larger volume
is affected by stress concentrations [71]. Especially large stress concentra-
tions develop at the edges or ends of anisotropic particles; according to
Riley et al. [29] stress concentration increases with increasing aspect ratio.
On the other hand, Nicolais and Nicodemo [92] claim that increased
aspect ratio leads to a decrease of stress concentration points.
Stress concentration and local stress distribution can be estimated by
the use of theoretical models or by finite element analysis [93-95]. The
interaction of stress fields of neighboring particles is very complicated and
changes with composition, so simplifying assumptions are used. The most
frequently used approach is the analysis of stresses around a particle
embedded in an infinite matrix, which was first applied by Goodier [87].
Goodier's approach is frequently used in stress analysis [71,93,94,96]. The
principal correlation of the analysis is:
VVu+(1-2v)Au=0 (1)

which must be solved under the following boundary conditions:


um(R) = uf(R) (2)

8 m(R)=8 f (R) (3)

where u is the deformation vector, 8 is the stress tensor and v is Poisson's


ratio. Superscripts m and f stand for matrix and filler, respectively.
14 Particulate filled polypropylene: structure and properties
Solution of Equation 1 under these conditions gives the distribution of
stresses around the particle. The radial stress component is shown in Figure
1.10 as a function of position on the surface of the particle. The value of (JrT
reaches a maximum at the pole and its value is almost twice that of the
external stress. Stress concentration in the tangential direction shows a
similar distribution, its value is, however, significantly lower. Nezbedova
and Davidovic [97], using finite element analysis, obtained very similar
stress distributions to that presented in Figure 1.10.

2.------r---,-----;----,---..,..----,

·-t····················r

0.5········.......... 1··.· ..••


!
or----+--;.---+---1----""~----1
Iwithoui therm~1 stress!
-0.5
I
············T············
.i ......... +;. . . . . . . .;;. . . .
-1+--~--;.--~--+--+--~
I
o 15 30 45 60 75 90
Polar angle, ()
Figure 1.10 Stress concentration in the radial stress component as a function of
polar angle. Effect of thermal stress. Rm = O.

Stress distribution around the particles is further modified by thermal


stresses induced by the different thermal expansion coefficients of the
components and the crystallization [84,85]. Nielsen [71] described the
thermally induced stresses by a simple correlation, while Vollenberg [93]
used an equation which yields thermal stresses changing also with composi-
tion, but this composition dependence is very weak (Figure 1.11). The
correlation most often used in heterogeneous polymer systems, including
PP blends and composites, was developed by Beck et al. [98], and gives
thermal stresses that are independent of composition

(4)

(5)
Stress analysis, micromechanical deformations 15
25.-------r------.-------r----~

i : j
20 ..........1 ............... ················f···············_···· ···.······t··.·.················

I !;
, I
·······························j····················I·························t····..·····.. ······..·········
! i
"'
E
,
iii 10 ··················_·············f·····
.J::.
I-

5 ................... .
. . . . .1!·································r····················.............!
,. . . . . . ..
!

o+-------+-------~-- __~~____~
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.11 Composition dependence of thermal stress calculated according to
Vollenberg [93] for glass bead filled PP composites.

where C( is the coefficient of thermal expansion, Ll T the temperature dif-


ference giving rise to the stresses, Gm the shear modulus of the matrix and
Rm the ratio of component shear moduli, i.e. Rm = GmjGf. Thermal stresses
decrease the radial stress component (Figure 1.10), but increase stress
concentration in the tangential direction [95].
Another factor which must be taken into account in the stress analysis is
the interaction of the components [88,99]. Interaction can vary in a wide
range, both in character and strength. Purely adhesive interactions can be
taken into account by Wu's equation [100]
(6)
where Wmf is the reversible work of adhesion and yd and yP are the dis-
persion and polar components of surface tension, respectively. Surface
treatment of a filler or reinforcement can lead to covalent bonding; in such
cases the quantitative prediction of the strength of interaction is difficult,
as it is in the case of other interaction mechanisms such as interdif-
fusion [83].
Under the effect of external load, inhomogeneity of local stress distribu-
tion initiates deformation at particular locations in the composite. Micro-
mechanical deformations playa crucial role in the deformation and failure
behavior of composites and especially in their impact behavior. The basic
deformation mechanisms of unmodified polymers are shear yielding and
crazing, and the dominant mechanism of a polymer is determined by its
16 Particulate filled polypropylene: structure and properties
inherent properties [101,102]. PP deforms mainly by shear yielding [96J,
although crazing was also observed in some cases, especially in its hetero-
geneous systems [70, 103].
The basic micromechanical deformation mechanisms are competitive
processes and the dominant one depends on material properties and load-
ing conditions. Many attempts have been made to define criteria for the
individual processes. The most commonly used condition for shear yielding
was defined by von Mises [96,104-106]:
(7)
where (1ym is the tensile yield stress of the matrix, and (11) (12 and (13 are the
principal stress components. Introducing stress components obtained from
stress analysis as described above into Equation 7, the position of yield
initiation and the magnitude of composite yield stress can be calculated. The
analysis shows that in the case of rigid particles shear yielding is initiated at
about 45° at the surface of the particle (Figure 1.12), a result corroborated
by experimental data, at least in glassy, amorphous polymers [102,107].
Direct visualization of the phenomenon in PP composites is much more
difficult.
More complicated and less unambiguous are the criteria for crazing. The
most frequently used conditions were given by Stern stein and Ongchin, who
expressed the criterion in terms of a stress bias, while a somewhat more
successful criterion was suggested by Oxborow and Bowden [96,104-106].

2.----,----.---.----,----,---,
I ! i
1.5 ····t.................'1"..... ············i···········
.............! ! i

I I
"." ....... , " , .. 1,
..· .... :..... .

without thermal stress


··················.. t····················!·············......!........ ··········1····················1··············

g 0.5 ................ L. . . . ~ I i I
~o 1
!
O+----+--~+r---+--~~r-~--~
u ,

-0.5 -f .................... '/. '··'·""··""",·" .. ,..·"·,,,.. "'t'·,

. .I
~ - 1 '

-1.5 ,~:::
iwith thermal stress
-2 ···r······ ,.,i....,...,' ·········1'······· ··········1' ········t······
!

- 2.5 +-----i------+-----+-----+-----;-----' I
o 15 30 45 80 75 90
Polar angle, (J

Figure 1.12 Position of yield initiation at the surface of the particle. Effect of
thermal stress. Von Mises condition (see Equation 7). Rm =0.
Stress analysis, micromechanical deformations 17
Stress components can be introduced into these criteria and conditions
of craze initiation can be calculated in the same way as for shear yielding.
In heterogeneous polymer systems and especially in composites these
basic deformation mechanisms are accompanied or replaced by another
mechanism, debonding. Debonding is especially important in PP com-
posites; because of the low polarity and consequently low surface free
energy of this polymer, interfacial adhesion is weak and separation of
the matrix-filler interface takes place under the effect of even moderate
load. Despite its importance, an accepted criterion for the initiation of
this mechanism does not exist. Taking into consideration the energy
balance for the creation of new surfaces Vollenberg and colleagues
[93, 108] developed a criterion, which, however, contains some assump-
tions. The correlation relates debonding stress to adhesion and to the size
of the particles.
Taking a similar approach, but without the assumptions, Puktmszky and
Voros [95] developed a model for the prediction of debonding stress. The
model assumes that the energy necessary to create new surfaces during
debonding is supplied by the change of elastic energy:

(8)

where AS represents the created new surfaces (Figure 1.13). The elastic
energy before debonding can be expressed as:

u o ="21 f mO
(Jik Elk dV+"21
0 f fO fO
(Jik Eik dV (9)
vm VI

Oebonding

before after

Figure 1.13 Creation of new surfaces in debonding.


18 Particulate filled polypropylene: structure and properties
which after debonding changes to:

1
UD = 2 f aik tik mD mD dV +21 f aik tik JD JD dV (10)
vrn VI

where ark, t& and ag, d are the stresses and deformations before and after
debonding.
After carrying out the necessary operations and including the effect of
thermal stresses the final correlation was obtained for the criteria of de-
bonding:

(11)

where aD and aT are debonding and thermal stresses, respectively, and C 1


and C 2 are constants containing component properties and geometrical
parameters. The correlation is essentially the same as the one derived by
Vollenberg [93].
In Figure 1.14 yield stress, which is proportional to the debonding stress
[109], is plotted against the square root of particle radius calculated from
the specific surface area of the filler. Specific surface area gives a more
accurate estimate of the average particle size than the value obtained from
sedimentation measurements, for reasons discussed in the previous section

28 ~----'-----r-----r----'-----.

26 ...... ,".........................,1.......... ·············l'!························r·············

I I ! i
,
24 . . . . . . . 1. . . . . . . . . . . . .1........... .............1.; ........................
!
! I.
!
I
I ......1
,
i
22 ···············f'························l···········

1
,
i
I
20 ·············t·························t·························t······..················

18 +------j-----i-------ir------T-----' I I
o 0.5 1.5 2 2.5
(Atpt/3)1/2(m -1/2)
Figure 1.14 Dependence of tensile yield stress of PP-CaC0 3 composites on the
particle size of the filler (see Equation 11). Filler content: 20 vol%.
Stress analysis, micromechanical deformations 19
and demonstrated by Figure 1.7. The figure proves the validity of the
treatment and the developed correlation.
Model calculations were also carried out to determine the value of de-
bonding stress for particulate filled PP composites. The values are listed in
Table 1.2. In the particle size range of real fillers, i.e. average particle size
1-3 ~m, the calculations predict that shear yielding is the dominating
deformation process, since the calculated (1D values are higher than the yield
stress of the matrix. Vollenberg, Heikens and Ladan [108] have carried out
similar calculations and drawn the conclusion that in the case of such fillers
no surface treatment is necessary to increase adhesion since debonding does
not take place. Volume strain measurements, however, have shown that
debonding takes place even in composites prepared with relatively small
CaC0 3 particles of 1.3 ~m diameter. The dominant character of dewetting
in such PP composites has also been proven by scanning electron micro-
scopic (SEM) studies (Figure 1.15) [109].

Table 1.2 Model calculations for the determination of debonding stress in


PP-CaC0 3 composites. Effect of particle size and relation to matrix yield stress

Particle size Specific surface area Debonding stress


(~) (m2jg) (MPa)

60.0 0.5 52.7


3.6 3.3 207.6
1.3 5.0 279.8
Matrix yield stress (MPa): 32.4 MPa

Both the analysis and the data of Table 1.2 indicate that yield stress of the
composite cannot exceed that of the matrix. Because of the effect of stress
concentration, local stresses must always be higher than external stress, thus
initiating local yielding. The data of Figure 1.16 prove, however, that
reinforcing can be achieved even with particulate fillers. These data show the
imperfection of the analysis, which stems from the simplifications and basic
assumptions, on the one hand, and from the effect of interfacial interactions,
on the other. The formation of a hard interlayer can lead to stronger
material. This has not been included in any of the theories to date, but can
be taken into account in finite element calculations, as shown by the results
of Nezbedova and Davidovic [97].
Although stress analysis and study of the criteria for micromechanical
deformation need further improvements they have shown the most import-
ant factors influencing local deformations. The role and importance of
these deformations in failure processes and especially in impact have been
proven many times. Although on a few occasions crazing has been detected
[70,110], in most cases dewetting proved to be the dominant deformation
20 Particulate filled polypropylene: structure and properties

Figure 1.15 Debonding in a PP-CaC0 3 composite deformed up to 2Ey (yield


strain) deformation. Average particle size: 1.31!m, filler content: lOvol%.

40~----~----~----~-----r-----'

I
i
Ii
"'
a...
:2
! ! i
·······················t·········_············t·····_·····_·········t········_··············

I I!
UJ

I
UJ
~
t)
"0 i, 1
Qi I ",
i ,
':;'
~
'iii
c:
20 ·························t······················~i

I I'"
(I)
I-

I I
I
10+-----~----~-----r-----r~--~
I
o 0.1 0.2 0.3 0.4 0.5
Volume fraction of filler (-)

Figure 1.16 Effect of filler particle size on the tensile yield stress of PP composites.
Particle diameter: (0) O.Q1l!m, (L"l.) 0.08I!m, (0) 3.3l!m, ('7) 58.0 I!m. (---) Theoretical
prediction calculated according to Nicolais and Narkis [201].
Interfacial interactions 21
process in particulate filled PP composites [111-114]. The main factor
in impact resistance seems to be the plastic deformation of the matrix
[79,88, 110]. Any of the above analyzed micromechanical processes, how-
ever, consumes energy and thus debonding and the related plastic deforma-
tion can also increase impact resistance, proven by the maximum in the
composition dependence of impact strength (Figure 1.17) observed by
several research groups [79,113,115,116]. The competitive character of
micromechanical deformation processes, their different composition depend-
ence, formation of a rigid interphase and interaction of particles at high filler
content lead to large variations of the composition dependence of impact
strength, which depends on particle and matrix characteristics, specimen
size, loading conditions, etc.

2.-----.----,----.------,

I
;

11.5 ·············-·········-r·······~··-··---·········{··-. __ ...-....................

---~----_J_------
.&:
C,
c:
~
1;)
ti : I
~
.s
.,.,
CD
.>
I
£ 0.5 : . I
...••................•....•.......;. .....•..............··· ..········1··········..··_··......·····_·t·_··_······..·· ................

I
I
!
o+----+---~--~---~
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.17 Composition dependence of the relative impact strength of PP-
CaC0 3 composites measured at 23°C: (0) ground marble, (6) precipitated CaC0 3
[116].

1.5 INTERFACIAL INTERACTIONS


Non-treated fillers and reinforcements have high energy surfaces [5]. Dur-
ing the almost exclusively used melt mixing procedure, polymer chains are
adsorbed onto the active sites of the filler surface. Adsorption of polymer
molecules leads to the development of a layer which has properties different
from those of the matrix polymer [40, 57, 99, 117]. Although the character,
thickness and properties of this interlayer or interphase is a much discussed
topic, its existence is now an accepted fact.
22 Particulate filled polypropylene: structure and properties
In semi-crystalline polymers matrix-filler interaction changes both the
structure and the amount of the crystalline phase. Because of the preferred
adsorption of large molecules, the dimensions of the crystalline units, which
depend on the molecular mass of the polymer, can change, usually decrease
[84]. Preferential adsorption of large molecules has been proved also by
gel permeation chromatographic (GPC) measurements after separation of
adsorbed and non-attached molecules of the matrix [118-121]. Decreased
mobility of the chains also affects the kinetics of crystallization. Kinetic
hindrance leads to the development of small, imperfect crystallites, forming a
crystalline phase of low heat of fusion [35,118,119].
Atomistic simulation of an atactic PP-graphite interface has shown that
the local structure of the polymer in the vicinity of the surface is different in
many ways from that of the corresponding bulk. Near the solid surface the
density profile of the polymer displays a local maximum, the backbone
bonds of the polymer chains develop considerable parallel orientation to the
surface [122]. This parallel orientation due to adsorption can be one of the
reasons for the trans crystallinity observed in the case of many anisotropic
filler particles.
Decreased mobility of adsorbed chains has been observed and proved
in many cases, both in the melt and in the solid state [90,121,123], and
changes in composite properties are very often explained by it [99, 121, 123].
Overall properties of the interphase, however, are not completely clear.
Based on model calculations, the formation of a soft interphase is claimed
[118], while in most cases the increased stiffness of the composite is
explained with the presence of a rigid interphase [55,90,99, 117]. The
contradiction obviously stems from two opposing effects. Imperfection of
the crystallites and decreased crystallinity of the interphase should lead
to lower modulus and strength and larger deformability. Adhesion and
hindered mobility of adsorbed polymer chains, on the other hand, decrease
deformability and increase the strength of the interlayer.
The thickness of the interphase is a similarly intriguing and contradictory
question. It depends on the type and strength of the interaction, and values
from 1 nm to several micrometers have been reported in the literature for the
most diverse systems [81, 119, 120, 122, 124, 125]. Since interphase thickness
is calculated or deduced indirectly from some measured quantities, it also
depends on the method of determination. Table 1.3 presents some data for
different particulate filled systems. The data indicate that interphase thick-
nesses determined from some mechanical properties are usually larger than
those deduced from theoretical calculations or from extraction of the filled
polymer [81,119,120,122,125-128]. The data supply further proof for the
adsorption of polymer molecules onto the filler surface and for the decreased
mobility of the chains. Thermodynamic considerations and extraction
experiments yield data that are not influenced by the extent of deformation.
In mechanical measurements, however, deformation of the material takes
I nterfacial interactions 23
Table 1.3 Interphase thickness in heterogeneous polymer systems determined
by different techniques

Matrix Method of Thickness


polymer' Filler determination (11m) Reference

HDPE Si0 2 Extraction 0.0036 [119J


HDPE Si0 2 Extraction 0.0036 [120J
PP Si0 2 Extraction 0.0041 [120J
PP Graphite Model calculation 0.001 [122J
PS Mica Dynamic mechanical 0.06 [125J
measurement
PMMA Glass Dynamic mechanical 1.4 [125J
measurement
PU Polymeric Dynamic mechanical 0.36--1.45 [127, 128J
measurement
PP CaC0 3 Young's modulus 0.012 [126J
PP CaC0 3 Yield stress 0.15 [126J
PP CaC0 3 Tensile strength 0.16 [126J

aHDPE, high density polyethylene; PS, polystyrene; PMMA, polymethylmethacrylate;


PU, polyurethane.

place in all cases. The specimen is deformed even during the determination
of modulus. With increasing deformations the role and effect of the immobi-
lized chain ends increase and the determined interphase thickness also
increases (see Table 1.3) [126].
Interphase, or more exactly the immobilized polymer layer, has the same
effect as increased filler content; very often a modulus increase is observed
and relative strength increases, always with increasing amount of interphase
[76,90, 129]. This increased apparent or effective filler content leads to a
decrease of the maximum amount of filler that can be introduced into the
polymer matrix [81,130]. Taking into account the spatial arrangement of
the particles, the maximum filling grade, i.e. maximum packing fraction
(q>j'ax) can be calculated. In the case of spherical particles it ranges from 0.52
to 0.74, while it can be as high as 0.91 for anisotropic fillers and reinforce-
ments [71,131]. In practice, however, incorporation of even much less filler
leads to unsurmountable difficulties or useless material. Table 1.4 shows the
maximum packing fraction of some CaC0 3 fillers calculated from expe-
rimental tensile yield stress data [81]. It is obvious that the value of q>j'ax
does not depend on packing since all the particles have practically the
same, more or less spherical geometry and similar relative size distribution.
Even the effect of aggregation can be excluded [71], since these fillers
have relatively low specific surface area and do not show any tendency for
aggregation.
Beside component properties and composition, properties of composites
depend on the size of the interface and the strength of the interaction. The
24 Particulate filled polypropylene: structure and properties
Table 1.4 Relation of the maximum amount of filler that can be incorporated into
the polymer (<praX ) and the particle characteristics of the filler in PP-CaC0 3
composites [81]

Particle size Specific surface area Maximum packing


(/lm) (m2/g) fraction (<p ax ) r
58.0 0.5 0.82
3.5 1.9 0.66
3.6 2.2 0.49
3.6 3.3 0.40
1.3 5.0 0.31

effects of these two factors are demonstrated by Figures 1.16 and 1.19,
respectively. Size of the interface is proportional to the specific surface area
of the filler, which is inversely proportional to particle size. In accordance
with the above proposed explanation of the relation of the effect of
immobilized polymer chains and the extent of deformation, modulus shows
only a very weak dependence on the specific surface area of the filler (Figure
1.18) [132]. Although the results of Figure 1.18 agree well with some
literature data [71, 129, 133], results showing considerable dependence of
modulus on the specific surface area of the filler can also be found in several
publications [55, 90, 117]. It is difficult to explain this contradiction. Many

3 ~------.------.------.-------.

2.5'" .......................+............................ I. . . . . . . . . . . . . . j............................ .


ca-
n...
~
(,
en
~
2 -----t-------r--------+
~
j .,1
. i
"0
o
E 1.5 .................................1"................................ 1•. ·········v··············t·····························....
'"Clco I !

o
ci5
-t---------)---------i------
!
0.5 -1----+-1---+---+---~
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.18 Effect of changing size of interface on the shear modulus of PP-
CaC0 3 composites. Particle diameter: (0) 0.01/lm, (6) 0.08/lm, (0) 3.3/lm, ('7)
58.0/lm.
I nterfacial interactions 25
possible reasons can be found, all of which need further study and verifica-
tion. It was shown previously, for example, that large filler particles detach
themselves from the matrix very easily, under the effect of small external
load. Debonding can even take place at the very low deformations of
modulus determination. Voids, however, decrease the modulus of the
composite dramatically. The presence of thermal stresses [71] or significant
interphase formation are other reasons that should be considered [117].
Properties measured at significantly larger deformations than modulus,
e.g. tensile yield stress or tensile strength, show much more pronounced
dependence on the specific surface area of the filler [132]. Figures 1.16 and
1.19 show that yield stress or strength values larger than the corresponding
value of the matrix can be achieved, i.e. even spherical fillers can reinforce
polymers [75, 76]. If adhesion is strong, yielding must be initiated at the
matrix value and no reinforcing would be possible [43, 134]. The reinforcing
effect of spherical particles can be explained only with the presence of an
interphase having properties somewhere between those of the polymer and
the filler [75, 76].
Interaction between PP and non-treated fillers is created by secondary,
van der Waals forces, which determine both the strength of the interaction
and the thickness of the interphase [124, 135]. Interaction strength changes
in a limited range since PP has very low surface free energy and the polar
component of surface tension is very close to zero [81]. Fillers, on the other

40~------~------r-------~-----.

10+-------+-------~------~----~
o 0.1 0.2 0.3 0.4
Volume fraction of filler ( - )
Figure 1.19 Dependence of tensile strength of PP-CaC0 3 composites on the size
of the interface (Ad. Particle diameter: (0) 0.Q1 ~m, (6) 0.08 ~m, (0) 3.3~, (v)
58.0~m.
26 Particulate filled polypropylene: structure and properties
hand, have high energy surfaces [5, 136J, but differences in the polar
component of their surface tension are larger than in the dispersive compo-
nent. Table 1.5 lists the surface tension of PP determined with contact angle
measurements, together with the same data for the three fillers most often
used in PP. For fillers surface tension and its components were determined
by the measurements of adsorption isotherms, either by inverse gas
chromatography (IGC) or by a gravimetric method [137].

Table 1.5 Dispersion (yd) and polar (yP) compo-


nents of the surface tension (y) of PP and selected
fillers

Surface tension (mJjm2)

Material yd yP Y

PP' 32.5 0.9 33.4


CaC0 3 b 54.5 153.4 207.9
Talc c 49.3 90.1 139.4
Silicac 94.7 163.0 257.7

a Contact angle measurements


b Measured by IGC
C Measured by a gravimetric technique

In composites adhesive interaction of the components is frequently


characterized by the reversible work of adhesion [81, 132, 138, 139]. In some
cases linear correlation has been found between some composite properties
and this thermodynamic quantity; in others, a more complicated relation-
ship was observed. Gutowski [138J has found a maximum in strength at
equal surface tensions of the components. Stress analysis and the criterion
for dewetting suggest that dewetting stress is proportional to the square root
of the work of adhesion [95, 108]. Figure 1.20 corroborates this prediction,
although the analysis does not take into account either the effect of
interacting stress fields of neighboring particles or the effect of the inter-
phase. Despite the apparently contradictory observations, work of adhesion
seems to be related to the macroscopic properties of PP composites, thus it
can be used for the prediction of interaction.
Maximum performance of a composite can be achieved only if wetting of
the filler or reinforcement by the polymer is perfect [28, 133, 138]. Treat-
ment of fillers with surfactants, e.g. stearic acid, is claimed to improve
wettability of the filler by PP due to changing polarity. According to Fox,
Hare and Zismann [140J, however, wetting of a high energy solid by a low
surface tension fluid is always complete, a condition which is completely
satisfied by PP and all inorganic fillers (see Table 1.5). Moreover, Huntsber-
ger [139J found that complete wetting, i.e. zero contact angle, is not even
Modification of matrix-filler interaction: Surface treatment 27
22 ,------,------,,------,------,

. .
CO" 20 -... -............,............:.,........ ·········,·································1··_·······............ .
a..
:::E !

i
·····························l·····················_·"I"_························r····················

. I i

14 +-----~------~------+-----~
7 8 9 10 11
(W mf ) 112(mJ/m2 )'12

Figure 1.20 Effect of strength of interaction (reversible work of adhesion, Equation


11) on the tensile yield stress of PP-CaC0 3 composites. qJr =0.3, Ar = 2.2 m 2 /g.

necessary for good adhesion. If wettability is characterized by the ther-


modynamic quantity:
Sm! =Y! -Ym -Ym! (12)
where Yr >Ym, wettability even decreases on surface treatment due to the
drastic decrease of the surface tension of the filler. The correlation is
demonstrated by Figure 1.21 where S m! is plotted against the surface cover-
age of a CaC0 3 filler with stearic acid [81]. The higher the value of S, the
better the wettability, and in the case of negative values definite contact angle
develops (partial wetting). Wetting, however, also has some kinetic condi-
tions, which depend on the viscosity of the polymer, mixing conditions and
particle characteristics, which are not always fulfilled during the preparation
of composites. Particle related problems (debonding, aggregation) and
insufficient homogenization usually create more problems than wetting.

1.6 MODIFICATION OF MATRIX-FILLER INTERACTION:


SURFACE TREATMENT
Changes in interfacial interactions can modify the mechanism of micro-
mechanical deformations, failure behaviour and, thus, the overall perform-
ance of composites. The obvious route to change properties and prepare
tailor-made composites is the modification of interactions. The possibilities
for changing the size of the interface are limited because of the role of large
28 Particulate filled polypropylene: structure and properties
50

40 ---_~--------'______L----- I I I
! I I
OJ
30 .... .................-......1................................1 . . . . . . . . _. __. . . ..1, ........_.....................
j ,
E
...,
g---
I i j
i i i
20 ...._......
i
.....-.-.....~.-...-.... -.... -...--..~I....-...-.....---....-..~I...--.-..-............-....
~ ! f i
:.c I! I
I'
i
:: 10 .................._...... +I ..............................;!................................1""i .............................
I
, J
i
o+------+------~~~-+----~

-10+------4------~------r-----~
o 25 50 75 100
Surface coverage (%)

Figure 1.21 Wettability of CaC0 3 by PP and its dependence on the surface


coverage of the filler with stearic acid.

particles in debonding, on the one hand, and the deteriorating influence of


aggregation, on the other [132]. The strength of the interaction, however,
can be changed in a wide range, if the type and amount of coupling agent is
properly chosen. Moreover, surface treatment can modify not only the
strength, but also the mechanism of interaction.
The type and mechanism of treatment must be chosen according to the
chemical and physical properties of the components, the conditions of the
intended application and the properties to be achieved [141-143]. Surface
modification can be classified in many ways. In this work, four arbitrary
categories are distinguished and discussed: (i) non-reactive treatment; (ii)
reactive treatment; (iii) formation of a polymer interlayer on the filler surface;
and (iv) development of an elastomer interlayer around the filler particles.
The oldest and most often used modification of fillers is the coverage of
their surface with a small molecular weight organic compound [35, 73, 81,
89]. Usually amphiphilic surfactants are used, which have one or more polar
groups and a long aliphatic chain. A typical example is the surface treatment
of CaC0 3 with stearic acid [73, 89, 144]. The principle of the treatment is
the preferential adsorption of the polar group of the surfactant onto the
surface of the filler, which is promoted to a large extent by the formation of
ionic bonds between stearic acid and the surface of CaC0 3 , but in other
cases hydrogen or even covalent bonds can also form. Surfactants diffuse to
the surface of the filler even from the polymer melt, which is further proof of
preferential adsorption [145].
Modification of matrix-filler interaction: Surface treatment 29
One of the crucial questions of non-reactive treatment, which is, however,
very often neglected, is the amount of surfactant to use. It depends on the
type of the interaction, the size of the treating molecule, its alignment to the
surface and some other factors.
The amount of bonded surfactant can be determined by simple tech-
niques. A dissolution technique proved to be very convenient for the
optimization of non-reactive surface treatment and also for the characteriz-
ation of the efficiency of the treating technology [137, 146]. Firstly, the
surface of the filler is covered with increasing amounts of surfactant, then the
non-bonded part is dissolved using a solvent. The technique is shown in
Figure 1.22, which presents a dissolution curve for a CaC0 3 filler. Surface
treatment is preferably carried out with the irreversibly bonded surfactant
(CIOO); at this composition the total amount of surfactant used for the
treatment is bonded to the filler surface. The filler can adsorb more
surfactant (crnax), but during compounding part of it may be removed from
the surface by dissolution or simply by shear, which may cause a deteriora-
tion in properties.

3 ~---,----.----.----~---,----,

j
2.5 ·········r····················.;. ·.;.····················l········

...
c:
III
t) o
~ 1.5 i
....................;.................. ·T····
:J
IJ)

"0

!· . . . . . . ··-l. . · . . . ·. ·. . t. · ·. ·. ··_··_··t...................j......................
Q)
• 1
"0
c:
o
al

........ : ......... _....... _.•!1................•....•l. ..


l
,,
0.5 ·,.··········-··········1·····
; i
; .i
o +-__ ~l_ _ _ _~_ _ _ _~_ _~_ _ _ _ +-__ ~

o 2 3 4 5 6
Total amount of surfactant (w%)
Figure 1.22 Dissolution curve used for the determination of surfactant adsorption
on a filler. CaC0 3 /stearic acid.

The specific surface area of the filler is an important factor which must be
taken into consideration during surface treatment. According to Figure 1.23
the amount of the irreversibly bonded surfactant depends linearly on it
[137]. Electron spectroscopy for chemical analysis (ESCA) studies carried
out on the surface of a CaC0 3 filler covered with stearic acid have shown
30 Particulate filled polypropylene: structure and properties
2.5 ..------.1---,.---..--.....,----.--------,
!

* +. . . . . . . . . +. . . . . . . .
2 ...............
1
'1""...··········l·················I·······

~ i i
j .5----r--r----l:--~
1

j----r---:-- r---r---r---
m o:-----+e---:----,---I"----l-
o 2 3 4 5 6
Specific surface area (m2jg)
Figure 1.23 Dependence of the amount of irreversibly bonded stearic acid on the
specific surface area of CaC0 3 fillers.

that ionic bonds form between the surfactant molecules and the filler surface
and that the stearic acid molecules are oriented vertically to the surface [137].
Further proof for the specific character of surface treatment is presented
in Figure 1.24. The fillers listed in Table 1.5 were treated with stearic acid.
The amount of stearic acid adsorbed on a unit surface differs significantly in
the case of CaC0 3 compared to the other fillers. Lack of specific interaction
in the form of ionic bond formation results in a significantly lower number
of irreversibly bonded molecules [137].
As a result of the treatment the surface free energy of the filler decreases
dramatically [81, 144, 146]. Lower surface tension means decreased inter-
facial tension and reversible work of adhesion as well, demonstrated also by
Figure 1.25, where the change of these quantities is plotted against the
surface coverage for a CaC0 3 filler treated with stearic acid [137]. Such
changes in the surface tension and in the thermodynamic quantities charac-
terizing interaction result in a decrease of both particle-particle and matrix-
particle interaction. One of the main goals, major reasons and benefits of
non-reactive surface treatment is this first effect, i.e. to change interaction
between the particles of fillers and reinforcements. Weaker interaction leads
to a dramatic decrease in aggregation, improved dispersion and homogene-
ity, easier processing, and better mechanical properties and appearance [73,
144, 147]. Improvement in the mechanical properties, particularly impact
strength, is often wrongly interpreted as the result of improved wetting and
interaction of the components.
Modification of matrix-filler interaction: Surface treatment 31

I
I
! l !
lO.75 ·····················t········· ·········f ······-············l··········=~--~-
Cl
x !
"*
.!
1
C ...i ......lS. ······t·················
!9 0.5
tJ !
~ i
:l
II)

"
Q)

"§ 0.25 ........ ·····r············ ·····1····


al

O..,..,L--+--j------t---t---t---'
o 2 3 4 5 6
Total amount of surfactant (w%)
Figure 1.24 Effect of specific interactions on the amount of adsorbed stearic acid.
Bonded stearic acid related to a unit surface for: (0) chalk, C",) marble, (\7) talc, (0)
silica.

150 .....---,.-----,----....,.------,

125 + ............................!...............................j ................................·j............·......·.......... 1

;:;' 100
E
...,
----
S
c 75 ..... ~.""~=. . . . . . . . . .
... i ................................i ............................. 1
o
'g
<t1
---Y'-~lr---~
~

C :: \ " ' ; ............................... + .............................."............................. 1


~~r---...lt.------.(>
o +----r----r----;----'
o 25 50 75 100
Surface coverage (%)

Figure 1.25 Effect of stearic acid surface treatment on the thermodynamic quanti ..
ties characterizing interaction in PP-CaC0 3 composites: (6) interfacial tension, (0)
work of adhesion.
32 Particulate filled polypropylene: structure and properties
As an effect of non-reactive treatment, not only particle-particle, but also
matrix-filler interaction decreases. The consequence of this change is de-
creased yield stress and strength as well as improved deformability [81, 144].
This is demonstrated by Figure 1.26, which shows the decrease of tensile
strength of PP/CaC0 3 composites with increasing surface coverage of the
filler. Adhesion and strong interaction, however, is not always necessary or
advantageous to prepare composites with the desired properties. As was
pointed out earlier, plastic deformation of the matrix is the main energy
absorbing process in impact, which decreases with increasing adhesion [79,
88, 110].

35 .-------~----~~-----,-------,

:
I
30 ..... . ...................i...............................f. .•••...•••••..
i
. . . . . . . . !.
1.: . . . . . . . . . . .

1 .

-'=
c,
c 25 ..............
I I
~
tl
~
.;;;
c

20-----r--
~

15+-----~------~------+-----~
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.26 Effect of interaction on the tensile strength of PP-CaC0 3 composites.
Stearic acid surface coverage: (0) 0%, (.6.) 25%, (0) 100%.

Reactive surface treatment assumes chemical reaction of the coupling


agent with both of the components. In the case of PP, because of its low
polarity and lack of reactive groups, it is difficult to find effective coupling
agents, although in some cases significant improvements in properties were
achieved with such treatment [80]. The great success of silanes in glass
reinforced thermosets has led to their application in other fields; they are
used, or at least have been experimented with, in all kinds of composites,
irrespective of the type, chemical composition or other characteristics of the
components.
Silane coupling agents are successfully used with fillers and reinforce-
ments that have reactive -OH groups on their surface, e.g. glass fibers, glass
flakes and beads, mica and other silicate fillers [40, 80, 141]. The use of
Modification of matrix-filler interaction: Surface treatment 33
silanes with fillers like CaC0 3 , Mg(OHh, wood flour, etc. has been tried,
but proved to be unsuccessful in most cases [142, 148]; sometimes contra-
dictory results were obtained even with glass and other siliceous fillers [149,
150]. The chemistry of silane modification has been extensively studied and
described elsewhere [67, 151, 152]. Model experiments have shown that a
multilayer film forms on the surface of the filler; the first layer is chemically
coupled to the surface, this is covered by crosslinked silane polymer, and the
outer layer is physisorbed silane [67, 152]. The matrix polymer may react
chemically with the coupling agent, but interdiffusion of the matrix and the
polymerized silane coupling agent, i.e. physical interaction, also takes place.
The complicated chemistry and surface structure of silane treatment was
demonstrated by Favis and colleagues, who studied the reaction of a silane
with mica (Figure 1.27) [153].

4.---------~----------~--------~----------,

x
-;0 3
I
·····························l························.........................!...
;
............... _···_··········1······································...........
j
!
(J
·E
!, ! i
··························............. 1.............................. Q .. : ··········1····· ...........................................
Q) 2
c: o !
.!!!
.iii I
-0
Q)
.............................1...........
I I
............... "1! ...............................................j................................................
.c
o
(/) .!" i. !
-0
« i j
i !
1
O~---------r--------~----------r---------~
i
o 50 100 150 200
Time (min)
Figure 1.27 Chemical reaction of a silane coupling agent with mica surface. Effect
of treatment time [153].

Although the chemistry of silane modification of reactive silica fillers is


well documented, much less is known about the interaction of silanes with
PP. In PP composites the most commonly used coupling agent is Z 6032
(Dow Corning), which is N-f3-(N-vinylbenzyl-amino) ethyl-y-aminopropyl
trimethoxy silane [40, 82, 148, 150]. Although one would expect that
coupling occurs through the reaction of free radicals, the chemistry is not
clear and further study is needed to reveal all of the factors influencing the
mechanism of interaction and the effectiveness of the treatment. Considering
the complexity of the chemistry involved, it is not surprising that the
amount of coupling agent and surface coverage have an optimum here too,
similarly to the surfactants in non-reactive surface treatment. This was
proved by Trotignon et al. [80], who found a strong increase and a
34 Particulate filled polypropylene: structure and properties
45 ~----~------~------~------.

!
42.5 .. ,.......................... .1 ............................. .J-.......................... ·····1·······························
I . .
I I
co
D... 40 ·······························t······················........ +I.............
,0
~
.!:
C,
i ; I
t: 37.5 ·······················_·1 ..............................
....~
~
'"
'(jj
j I
t: 35 ·..······i·······························i································1································
CD

··--··~---·-··-j-··-·--··-·i·
I-

32.5

30 +----+---+----f----l
o 0.5 1.5 2
Amount of silane (%)
Figure 1.28 Maximum effectiveness of silane treatment of mica in PP composites
[80].

maximum in the strength of PP-mica composites as a function of the silane


used (Figure 1.28). Optimization of the type and amount of coupling agent is
also crucial in reactive treatment, and although 'proprietary' treatments may
lead to some improvement in properties, they may not be optimal or cost
effective [154, 155].
Improper choice of coupling agent may lead to insufficient or even
deteriorating effects. In some cases hardly any change in the properties is
observed, or the effect can be clearly attributed to the decrease of surface
tension due to the coverage of the filler surface by an organic substance, i.e.
non-reactive treatment [90, 156].
Coverage of the filler surface with a polymer layer which is capable of
interdiffusion with the matrix proved to be very effective both in stress
transfer and in forming a thick, diffuse interphase with acceptable deforma-
bility. Because of its increased polarity, in some cases reactivity, maleic
anhydride or acrylic acid modified PP adsorbs to the surface of most polar
fillers even from the melt. Often a very small amount of modified PP is
sufficient to achieve significant improvement in stress transfer [157]. The
importance of interphase thickness was clearly demonstrated by the experi-
ments of Felix and Gatenholm [158], who introduced maleic anhydride
modified PP (MA-PP) into PP-cellulose composites and achieved an
improvement in yield stress proportional to the molecular mass of MA-PP
(Figure 1.29). Model calculations have shown that improvement of proper-
ties has an upper limit, and a plateau is reached at a certain, not too high,
Homogenization 35
45~----~~-----r------.------.

40 .. ········· ..··· ..·.... ·1......···..·....··..·..·..··.., ....

.,
0..
~ 35 •••••••••••••••••••••••••••••••••1- ••
i

~
'iii 25 ......
~ r-~~~r=~~~::::::~~::==
20 ·..·.... ·· ..······............../·..···· ..·· ....·...... ·· .... ··..t. · ·. · · . . . . . . . . . . .:. . . . . . . . . . .

15+-------r-----~------~-------"
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.29 Effect of interphase thickness on the tensile yield stress of PP/cellulose
composites. Molecular weight (Mw) of MA-PP used for modification: (0) no
treatment, (6) 350, (\7) 4500, (0) 3.9 x 104 [158].

molecular mass. The maximum effect of modified PP was found with fillers
of high energy surfaces [82, 149, 159, 160] or with those capable of specific
interactions, e.g. ionic bond with CaC0 3 [123, 158, 161] or chemical
reaction with wood flour, kraft lignin or cellulose [133, 142, 162].

1.7 HOMOGENIZA nON


In particulate-filled polymers the structure of composites is determined by
component properties, especially by particle characteristics and anisotropy,
as well as by processing technique and conditions. The most important
structure-related factors are aggregation, distribution of composition
throughout the product and orientation of anisotropic particles. All three
influence the properties of PP composites and must be controlled in order to
prepare products of high quality.
Interaction of particles, i.e. aggregation, is one of the major issues during
the preparation and processing of particulate-filled PP. Particle-particle
interactions result in inhomogeneous distribution of the filler, processing
problems, poor appearance and inferior properties.
Ess and Hornsby [163] have exhaustively analyzed aggregation and
homogenization of filler particles in a polymer matrix and listed a great
number of parameters influencing particle-particle interactions, such as
electrostatic charges, liquid bridges, etc. The two major factors, however,
36 Particulate filled polypropylene: structure and properties
which determine the aggregation tendency of particulate fillers, are particle
size and surface free energy. The analysis of the adhesive forces between two
bodies has demonstrated that the force acting between them depends on the
curvature of the contacting surfaces and on interaction [164, 165]. Assuming
particles of identical size and relating the force to a unit surface, specific
adhesion force can be determined [166]:

Wmf
(J
a
=K 1 - -
R (13)

Adhesion increases steeply with decreasing particle size, and Figure 1.30
indicates that it is in the same range as shear forces developing during
processing [167, 168]. The results shown in Figure 1.30 predict strong
aggregation of a filler with small particle size (e.g. precipitated CaC0 3 ,
aerosil), which is corroborated by Figure 1.31. In Figure 1.31 a polarization
micrograph of a PP-CaC0 3 composite is shown. The average particle size
of the filler is 0.08 !lm. Figure 1.30 also indicates that an increase of shear
stress shifts the appearance of aggregation toward smaller particle sizes; in
injection molding improved homogeneity and a decreased number of
aggregates are expected.
Determination of aggregation has great importance in quality control and
development of particulate-filled polymers. Different techniques can be used
for the estimation of the agglomeration tendency of particulate fillers.

200.-------~----~------~------,
!,i
~
150 ···_··············_···r·························l··········_················l···············_··········

I I
to
Q.

~ Injection molding
c: ----~--------~-------,--------
100 .............................~ .....-..................-.....~ ................................i ................................
!
0
'(jj
CI>
..c: i !
«
-0

50

O+-------r-----~-------+------~
o 2.5 5 7.5 10
Particle size (Ilm)
Figure 1.30 Dependence of adhesion between particles (aggregation tendency) on
particle size. Relation to shear stress in processing operations.
Homogenization 37

Figure 1.31 Aggregates in a PP-CaC0 3 composite. Polarization light microscopy.


Particle size: 0.08 11m.

Suetsugu and White [73] used a simple sedimentation method to character-


ize particle-particle interactions of CaC0 3 fillers with different particle sizes.
Results of this technique correlated well with measurements carried out with
polarization light microscopy on compression molded films. Suetsugu et al.
also developed a small angle light scattering (SALS) technique [72], which
quantitatively characterized aggregation in PP-CaC0 3 composites. Light
microscopy combined with automatic image analysis was used by Ess and
Hornsby [163, 168] for the quantitative characterization of aggregation,
while Svehlova and Poloucek [74] simply used the total area of aggregates
larger than 10 Ilm for this purpose. This simple quantity gave very good
correlation with some mechanical properties of PP-CaC0 3 composites.
Although some other techniques, e.g. optothermal measurements [86, 169],
determination of particle size distribution using different methods (see
Figure 1.7) and crystallization experiments (Figure 1.4) can also be used,
polarization or reflection light microscopy, image analysis, and statistical
evaluation seem to be the best techniques for the quantitative characteriz-
ation of particle aggregation in particulate-filled PP.
The effect of aggregation on composite properties is clearly detrimental.
Numerous reports emphasize this fact together with the importance of the
best possible homogeneity [29, 89]. Miyata, Imahashi and Aabuki [59]
prepared flame retarded composites by incorporating Mg(OHh fillers into
PP. The large specific surface area of the fillers indicates a significant
fraction of small particles and strong aggregation. Indeed, composites of
acceptable properties could not be prepared form the fillers without surface
treatment. On the basis of these results Miyata, Imahashi and Aabuki
38 Particulate filled polypropylene: structure and properties
concluded that fillers with fewer aggregates result in composites with better
properties. Svehlova and Poloucek [74J correlated the area occupied by
filler aggregates to impact strength and found a very strong correlation
between the two quantities (Figure 1.32). Increasing aggregation leads to a
drastic decrease of impact resistance of PP/CaC0 3 composites. Aggregation
is also detrimental in the case of anisotropic particles. Riley et al. [29J have
found that aggregates act as flaws in composites. All of these results
emphasize the importance of proper homogenization.

350

300
E
-,
----
~ 250 .l
.s::
0,
c
~
200
\ '\.
1ii
t5
co
c.
o~
'-..0

'"
.§: 150
~
'iii
cQ)
I-
100 ~ .......................
,
,
j
50
o 2 3 4 5
Relative area of agglomerates (%)
Figure 1.32 Correlation of impact resistance and aggregation [74].

Equation 13 offers possibilities for the improvement of homogeneity and


decrease of aggregation. Both factors of Equation 13 - particle size and
interaction - can be modified. Although fillers with smaller particle size
(larger specific surface area) give better reinforcing effect, larger debonding
force and sometimes even improved impact strength, an optimum must be
found between the beneficial influence on mechanical properties and the
detrimental effect on aggregation. Modification of interaction is clearly one
of the easiest and most convenient techniques (also the most frequently used)
to decrease particle-particle interaction and thus aggregation [59, 73, 147].
The approximately 210 mJ/m2 surface tension of a CaC0 3 filler can be
decreased to 40-50 mJ/m2 by stearic acid treatment, which leads to a
significant decrease of interfacial tension and work of adhesion (Figure 1.25).
As a result, particle interaction will decrease and lead to better dispersion
(Figure 1.33). Finally, Equation 13 also implies that shear forces exceeding
particle-particle adhesion separate the particles. This is supported by the
Homogenization 39
200~----~------~----~----~

1 50 -------+----------t---------+----------

I
·1·····················..·..··_·:········..··············........,...............................

o+-------~----~~-----+-------"
o 2.5 5 7.5 10
Stearic acid (w%)
Figure 1.33 Decrease in the number of aggregates in PP composites as an effect of
stearic acid surface treatment of wood flour [147].

observation of Bigg [170], who found that increased homogenization time


does not improve homogeneity, but increased shear and decreased interac-
tion does. The effect is shown in Figure 1.34, where composition dependence
of the tensile yield stress of compression and injection molded PP-CaC0 3
composites is presented. Although the samples contain the same filler, the
strength of the injection molded samples is clearly superior, due to the better
homogeneity.
A special form of heterogeneity is the processing-related segregation of
particles. Changes in local composition were observed in most processing
operations, but mainly in injection molding [171]. As a result of in-
homogeneous velocity distribution during flow, particle content changes in
the cross-section of the product; mass flow of filler is directed towards the
center of the flow path. Although segregation of filler particles during
processing is well documented [171, 172], several studies have shown that a
long flow path and large particle size are necessary to observe significant
composition differences in a product due to segregation.
As extensively discussed earlier, particle characteristics playa crucial role
in the determination of composite properties. Particle dimensions can,
however, significantly change during processing. Change of fiber length and
length distribution is an intensively studied question in short fibre reinforced
composites [173]. It is of similar importance, however, in composites filled
with anisotropic, plate-like filler particles, such as mica and talc. Cleavage of
these fillers is relatively easy and considerable delamination takes place
40 Particulate filled polypropylene: structure and properties

40 ~----~-------r------'------'

• •
Ij
. ························-T···························t ·····························r························...

.::
C,
c:
!!!
1ii
~ I
.~ 20 ;t;:;;;;;;;:l~~:--_6._..."'l
.. r:. . . . . . . . . . . . . i . . . ._............. .
I-

10+-----~------~------+-----~
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.34 Effect of processing technology (aggregation) on the compositIOn
dependence of tensile strength of PP-CaC0 3 composites: (6.) injection molded,
(0) compression molded.

during processing, especially in injection molding, where shear stresses can


be very high [76, 82, 150]. Delamination of mica particles was shown to
improve most properties, but decreasing particle size leads to aggregation,
with the associated disadvantages [77].
Another processing-induced structural phenomenon is the orientation of
anisotropic particles. The phenomenon and the resulting structure are
similar in both glass fiber reinforced and particulate-filled composites.
Plate-like, planar reinforcements, however, have some advantages over
fibers; orientation dependent shrinkage of particulate-filled composites is
significantly smaller than that of the fiber reinforced ones [5]. Orientation
and orientation distribution influence property distribution and the overall
performance of the product [174].
Orientation distribution of fibers and anisotropic particles is determined
by the flow pattern and shear forces developed during processing [175].
Orientation of plate-like particles is observed both in extrusion [176, 177J
and in injection molding; moreover, even the mild shearing conditions of
compression molding induce significant orientation [177]. Orientation dis-
tribution of fibers is usually determined using micrographs taken from
polished surfaces. Automatic image analysis greatly facilitates quantitative
characterization. Although the same techniques can also be used in the case
of particulate fillers [57, 178J, their planar geometry and the different optical
characteristics of the composites make the determination difficult.
Homogenization 41
A simple technique was developed by Rockenbauer et ai. [177]. They used
electron spin resonance (ESR) spectroscopy for the measurement of relative
orientation of filler particles. Practically all mineral fillers contain impurities
and defect centers in their crystalline structure. The orientation of magnetic
axis and the crystal planes are in unambiguous correlation, thus the
orientation of filler particles can be quantitatively characterized by this
method. In Figure 1.35 ESR spectra of PP composites containing oriented
talc particles and that of the randomly oriented talc powder are shown. The
degree of orientation can be characterized by the relative amplitude of the
two peaks observed on the spectrum of the sample with parallel orientation.

1\ I ~I
························I···········~~~~~r.·······
i I
Q)
"tJ
.~
0.
E
ro
IX:
en
LIJ

3100 3125 3150 3175 3200


Magnetic field (G)

Figure 1.35 Orientation dependence of the first hyperfine line of the Mn 2 +


impurity center of talc.

In Figure 1.36 composition dependence of average particle orientation


is plotted against composition for compression molded PP-talc compos-
ites. The measured orientation number is an average of oriented and non-
oriented regions (Figure l.37a and b). Local orientation of filler particles
is determined by local flow and shear conditions; parallel orientation was
observed at the wall, while more random distribution was seen in the
middle of the compression molded plates. Average orientation shows
significant composition dependence. The strong composition dependence
of orientation is further proof for particle-particle interaction, which re-
sults in hindered rotation. Detailed study of the orientation of injec-
tion molded parts has yielded similar results. The long axis of talc
particles was preferentially oriented into the flow direction; parallel
42 Particulate filled polypropylene: structure and properties

i -----,
0.5 .,.-------.----.,------r
I

I'
Q)
'0
0.4 ------t---r-----J------
I i
.~
C. ~
E
til
a:
0.3 ·. · . . ·. . . . . . . ···. t. ·. . ·. . · . ·. ·. ·'i' . ·........· . ·........·~.i..·........ ·
en 1
UJ !
Q)
>
.;:;

£
til

0.2 ·. . ·. . . . ·. . . ·. ·. . ·+·. . . . . ·. . . . . . . . ·j·......·................·......r ............................


!

0.1 +-------~-----4-------+----~
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.36 Amplitude ratio of the orthogonal components of the Mn2+ hyper ..
fine lines of talc (average orientation) in PP-talc composites as a function of the
composition.

(a) (b)

Figure 1.37 Orientation of talc particles in a compression molded PP composite.


Filler content: 10 vol % (a) particles oriented in the plane of the compression
molded plate, (b) non-oriented particles.

orientation of talc particles occurred near the wall, while in the center of
the part more random orientation was observed.
Average orientation of the particles relative to the direction of the
external load significantly influences the properties. Increasing alignment
Composition dependence of properties: Prediction 43
results in increased reinforcement, larger modulus, and higher stress and
impact strength, as demonstrated by the data of Mittal, Gupta and Sharma
[179] (Figure 1.38) on glass fibre reinforced PP. Similar observations were
also made on anisotropic particle-filled PP; increased orientation of filler
particles improved mechanical properties in talc, mica and other anisotropic
filler containing composites [57, 79, 176].

801~----'------.-----r----~-----.
1. i
I !
! i
70 . . . .!. . _. . . . . . . . . . +. . . ··················t··························t········
! ! ! 1
..................

-;a i . ,1••

i
I

i
II... 60 .~ ....................-....!......................-... t·······~··············-·t····
,I '
~

. . . . . . . . . . !................. '1" · ..·..1·......· ....·....·..

__ __+_____; ___+_0-_:
30 .. ·.... ·....·.. · .... II. . . . ·. . . . ·. . l· . ·. _. ·. . . ·;. . . . . ·. . ·. ·f· ............. .
1

20+-----+-----+-----+-----+---~
o 10 20 30 40 50
Fiber alignment (a)
Figure 1.38 Effect of fibre orientation (alignment to external stress) on the
strength of short glass fiber reinforced composites [179].

The orientation of anisotropic filler particles has an especially pro-


nounced effect on the strength of injection molded parts containing weld
lines.* Fountain flow in the mold leads to orientation of particles parallel
to the melt front and thus to the weld line, resulting in decreased weld line
strength [180-182]. Increasing particle size and filler content result in a
decrease of weld line strength [180, 183]. Improvement in weld line
strength can be achieved by changing particle characteristics (size, treat-
ment, aspect ratio) [180] and mold construction [181].

1.8 COMPOSITION DEPENDENCE OF PROPERTIES:


PREDICTION
Properties of particulate-filled thermoplastics depend strongly on composi-
ton. Knowledge of this dependence is important for both theoretical and

*Treated also in Chapter 7 of Volume 1.


44 Particulate filled polypropylene: structure and properties
practical aspects. The desired effect (stiffness, price decrease, functional
properties) increase with filler content, thus introduction of the maximum
possible amount of filler is one of the primary goals, both in filling and
reinforcing. Increasing filler content, however, leads to larger amount of
demobilized polymer in the interphase and structural effects, mostly to
particle-particle interactions. These latter, as we have already seen, are
usually accompanied by the deterioration of properties, i.e. the filler content
of the composite obviously has an upper limit.
The introduction of fillers or reinforcements changes practically all
properties of the polymer, including its rheological characteristics. Viscosity
usually increases with filler content, while melt elasticity decreases [184,
185]. These changes depend very much on the particle characteristics of the
filler although unambiguous relations are not known. Matrix-filler interac-
tion depends on adhesion or 'affinity' of the components and has the same
effect as increasing filler content [186].
Surprisingly enough, composition dependence of rheological properties is
modeled in only a few cases. The majority of the models used for disperse
systems are derived from Einstein's equation developed for the suspension of
spherical particles. The original equation is valid only at infinite dilution, or at
least very low (1-2%) concentrations [187] and in real composites modifica-
tion of the equation is necessary; additional parameters are introduced, most
often in the form of a power series [185, 187]. The Mooney equation represents
another approach which contains adjustable parameters accommodating
both the effect of interaction and particle anisotropy [71, 187, 188]. This
equation has also been used for the study of the viscosity of PP composites
[188-190], but the rather limited interest in this field indicates that particU-
late-filled PP often can be processed without serious difficulties.
At high filler content interaction of the particles takes place, resulting in
the formation of some higher order structures, and a yield stress appears.
More attention is paid to the investigation of such phenomena. Casson plots
[71, 191] or modified Casson plots [192] facilitate the detection of aggrega-
tion. Other structure related phenomena in melt flow are the orientation of
anisotropic particles [191] and segregation [172]. Both were observed in PP
composites, but no model equation or theoretical treatment was used to
describe them.
Stiffness is one of the basic properties of composites; the aim of filling is
very often to increase it [39, 157]. The rate of increase depends mostly on
the type and especially on the anisotropy of the filler. Higher anisotropy
leads to increased modulus, demonstrated also by Figure 1.8 [5]. Often
linear dependence of modulus on composition is observed, although theor-
etical models predict nonlinear composition dependence. One important
factor, however, is frequently neglected - including in Figure 1.8, taken from
the literature [5] - the orientation distribution of anisotropic particles.
Orientation, as discussed earlier, has a crucial effect on the value and
Composition dependence of properties: Prediction 45
composition dependence of modulus; moreover, it also changes with compo-
sition (see Figure 1.36).
Modulus is not only the most frequently measured, but also the most
often modeled composite property. A great number of models exist which
give some kind of a prediction for the composition dependence of this
characteristic or at least some bounds for its value. The abundance of
models is relatively easy to explain: modulus is determined at very low
deformations and thus linear viscoelasticity theory can be used in the
development of model equations. The large amount of accessible data also
helps, both in the development and verification of models.
Model equations developed for heterogeneous polymer systems can be
classified in different ways [193-195]. Leaving out of consideration the
completely empirical correlations, four groups are distinguished here:

1. Phenomenological equations which are similar to the spring and dash pot
models used for the description of viscoelastic properties of polymers.
2. Bounds, which are usually exact mathematical solutions that do not
contain any or contain only very limited assumptions about the structure
of the composite.
3. Self-consistent models. The mechanical response of a composite structure
is calculated in which the simple or multiple embedding of the dispersed
particle into the continuous phase is assumed. A well-known model of
this type, also frequently used in PP composites, is the Kerner equation
[193-195]. Although it was much criticized because of the incorrect
elastic solutions used [196], the model gained wide use and acceptance.
4. Semiempirical models. In spite of the effort of the self-consistent models
to take into account the influence of microstructure, they very often fail
to predict correctly the composition dependence of composite modulus.
To improve their performance additional, adjustable parameters are
introduced. The most frequently applied equation of this type is the
Nielsen (also called Lewis-Nielsen or modified Kerner) model [71, 197].

G=G m l+ABcpf (14)


I-B'¥CPf

(15)

(16)

The equation contains two structure-related or adjustable parameters. The


two parameters, however, are not very well defined. A can be related to filler
anisotropy, through the relation A = kE -1, where kE is Einstein's coefficient.
Although the correlation is sometimes used, it has not really been investigated.
46 Particulate filled polypropylene: structure and properties
'I' depends on the maximum packing fraction. cp}"ax is related to anisotropy,
but it is influenced also by the interphase, which was not taken into consider-
ation in the original treatment [71]. Its experimental determination is difficult;
moreover, McGee and McCullogh suggested a different form for 'I' [193].
Despite these uncertainties the correlation is quite frequently used in PP
composites for the prediction of the composition dependence of modulus. In
some cases merely the existence of a good fit is established, in others
conclusions are drawn concerning interactions or observations are made
about the structure of the composite. Attention must be drawn, however, to
the problems of the application of these equations or any other theoretical
model. Figure 1.39 presents the composition dependence of Young's
modulus for PPjCaC0 3 composites. Particle size varies in a wide range and
a filler treated with stearic acid is also included, i.e. both size of the interface
and strength of the interaction change [132]. The influence of changing
interaction on modulus is small; changes are in the same range as those
resulting from using different filler moduli in the calculations (Table 1.6). It
is clear that uncertainty in input values has the same effect on the prediction
as changing interaction on measured data.
Similarly, Figure 1.40 presents composite moduli for PP composites
containing fillers with different degrees of anisotropy, together with predic-

5.-------r-----~-------.------~
I
i

!
j I I

., 4 .................................-r .................._.............t-............................... , .............;;:;.:~~


Q.. i ! /,/
~ I I' I /,,//,-
i "" "
UJ
::J i!! I /i~/ ... "
"'8
:;
3 .................................;.................................~-....... ..
1 " .... 1 ,-
-;;··~··;,.·;;r·························-·····

E 1 /~~ / / j
i . . j~~~;/
"V
1
~ ,::: ~~ ~r
~ ~ ~. ~ (·~··~·:.··~···········l··· . . . . . . . . . . ····T _. _..................
I I I
! I
I
!
i
1+-------+-------+-------+-----~

o 0.1 0.2 0.3 0.4


Volume fraction of filler (-)
Figure 1.39 Composition dependence of the Young's modulus of composites
prepared with CaC0 3 fillers of different specific surface area and treatment. Ar:
(0) 0.5 m 2/g, (Ll.) 3.3 m 2/g, (0) 5.0m 2/g, (v) 3.3 m 2/g, treated. Comparison to predic-
tiOn (Equations 14-16) calculated with different modulus values for the filler taken
from the literature, i.e. 19.5 [198], 35 [93] and 47 [200], GPa in increasing order.
Composition dependence of properties: Prediction 47
Table 1.6 Young's modulus and Poisson's ratio values reported
in the literature for CaC0 3

Young's modulus
(GPa) Poisson's ratio Reference

19.5 [198]
26.0 0.27 [199]
35.0 0.20 [93], p. 17
47.0 0.30 [200]
50.0 0.25 [93], p. 52

3.------,-----,-----,-----,

2.5 ········..······..····..··+..··········..········ ..·1· ..··········.. ··········t··········


'"
0...
~
~ 2 ·t············
! . /
:::l
"S ~ /
/! "
-c ~ ....
o .,/

."""";" j'::,::-"1------
: " ......... 1, . .
~ 1.5············ .. ·
Cl

o'"
U5
................_....-i ........_.
i

I
0.5 +------ir----+----+-----..J !
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.40 Effect of anisotropy on composite shear modulus. (L,,) sphere, (0) short
carbon fibre. Prediction (Equations 14-16) calculated with different cpf"'x values of
0.91 (parallel hexagonal packing), 0.64 (random close packing), and 0.52 (simple
cubic packing), in increasing order.

tions calculated with 'II parameters changing in the possible range of


maximum packing fraction values (see [71], chapter 7, table 1). The effect is
obvious. Similar calculations can also be made for parameter A. A better fit of
the predictions can be made in Figures 1.39 and 1.40 with a different choice
and combination of the parameters E r, Gr, cpf'ax and A. Figures 1.39 and 1.40
unambiguously prove that modeling modulus data, and the conclusions
concerning interaction, structure, etc. must be treated with caution.
While the fact that modulus is determined at very low deformation
simplifies both measurements and modeling, yield properties are measured
at considerable deformations, which leads to more controversy. As an effect
48 Particulate filled polypropylene: structure and properties
of filling yield strain almost invariably decreases, due to the decreased
deformability of a rigid interphase [49, 69, 156]. Composition dependence of
yield stress usually shows a similar tendency [48, 133, 148], which is often
erroneously generalized and identified with the yield behaviour of particu-
late-filled composites. Depending on particle size, interaction or orientation
of anisotropic particles, occasionally also true reinforcing, can be observed
[75, 83, 113, 133]; this is demonstrated by Figures 1.9 and 1.16.
Composition dependence of yield strain is described under the assump-
tion that at the considerable deformations (5-10%) of yielding only the
polymer matrix deforms and the rigid filler does not. This strain magnifica-
tion of the matrix increases with increasing filler content. One of the
equations based on this principle was derived by Nielsen [71]:
f y = f ym (l_<p}l3) (17)
where fy and fym are the yield strain of the composite and matrix,
respectively. Several similar equations were developed and Equation 17 is
sometimes used in a modified form [69]. Occasionally the same equation is
used for the prediction of the composition dependence of elongation at
break [71, 90]. Although the equations describe the general tendency of the
composition dependence of yield strain, deviations from the prediction are
frequent, thus yield strain is rarely modeled with these equations.
More attempts have been made at predicting and analyzing yield stress.
The most frequently applied correlation was developed by Nicolais and
Narkis [201], although the equation of Ishai and Cohen [202] is practically
the same. Nicolais and Narkis assumed that filling decreases the effective
cross-section of the matrix which carries the load during deformation.
Assuming a certain arrangement of the particles they calculated this cross-
section and from that the dependence of yield stress on composition:
O"y =O"ym(l-1.21 <p;J3) (18)
where O"y and O"ym are the composite and matrix yield stress, respectively.
Their derivation, however, leads to a matrix cross-section assuming zero
value at <Pc < 1, which is, of course, not true. Even the argument that losing
matrix continuity results in zero yield stress is not valid, since samples break
without yielding at much lower filler contents than <pr ax ; in this range the
correlation cannot be applied at all. The model does not take into consider-
ation the effect of stress concentration, interaction or any other factors
influencing yield stress.
Because of these simplifications, deviations from the prediction occur very
often. They are usually taken into account by modification of the model if
the effect of different arrangement of the particles [71], stress concentration
[90, 92], etc. is included. Dependence on the latter factor is taken into
account by a parameter, which is substituted into the place of the con-
stant, 1.21. Experimental data, however, indicate that deviation from the
Composition dependence of properties: Prediction 49
prediction is almost exclusively positive (Figure 1.16), which cannot be
explained by the effect of stress concentration. The magnitude of local stress
maximum does not depend on particle size, while yield stress does, also
demonstrated by Figure 1.16. Sometimes the same, adjustable parameter is
related to matrix-filler interaction [69], i.e. its physical meaning is not
completely clear either.
Pukanszky, Turcsanyi and Tiid6s [75] have taken an approach similar to
that of Nicolais and Narkis. However, they used a different expression to
take into account the effect of decreasing load-bearing cross-section [131]
and extended the model to account for the effect of interfacial interaction:
1-q>J
uy=u Yml 25 exp(Bayq>J) (19)
+ . q>J
where Bay is a parameter related to the relative load-bearing capacity of the
components, i.e. to interaction. A detailed analysis has shown that Bay
accounts for changes in both the size of the interface and the strength of the
interaction through the expression:

(20)

where Ar and Pr are the specific surface area and density of the filler, and I
and Uyi are the thickness of the interphase and its yield stress, respectively.
The correlation proved to be valid in most particulate filled systems, but
its validity was mostly checked on PP composites [75, 81, 132]. Rearrange-
ment of Equation 19 eliminates the effect of changing matrix cross-section
and if the natural logarithm of the reduced tensile strength is plotted against
the volume fraction of the filler, straight lines should be obtained. The
validity of the treatment is proved by Figure 1.41, where the data of Figure
1.16 are plotted in a linearized form. The change in the slope of the straight
line corresponds to the effect of the increasing size of the interface (Ad with
decreasing particle size, i.e. to increasing interaction. Changes in the par-
ameter Bay express quantitatively also the influence of different surface
treatments (Figure 1.42); the increased adhesion due to the interdiffusion of
the MA-PP layer and the matrix polymer is clearly seen in the figure.
Composition dependence of ultimate tensile properties, i.e. tensile strength
and elongation at break, is very similar to that of the yield characteristics.
Both elongation and strength decrease with increasing filler content [59,
203], although occasionally reinforcing effect can also be observed (Figure
1.19). Even fewer models are known which predict ultimate properties than
for yield stress or yield strain. The models developed for yield properties can
be and are also applied for strength and elongation, if deformation is low at
all compositions, e.g. particulate filled epoxy [204]. PP composites, how-
ever, can be deformed to several hundred percent of their original length at
low filler content « 5 vol %), but their elongation is very small, at filler
50 Particulate filled polypropylene: structure and properties

0.8 -r----...,---....---------~

I
0.6 . . . . . . . . . . . . .1. . . . . .
!
!

0~--~---+---4---~-~
o 0.1 0.2 0.3 0.4 0.5
Volume fraction of filler (-)

Figure 1.41 Relative tensile yield stress of PP-CaC0 3 composites of Figure 1.16
plotted against composition in a linearized form. Particle diameter: (0) 0.01 11m,
(.6) 0.08 11m, (0) 3.3 11m, ('l) 58.0 11m.

0.5 -r-----,1---..----...,..---....----,

125 ----+---I----t---
i

!075---=I~]
8=3.40!

'>' ! '
.,
CD
.>
co 0.5
!

OtF---i---i----i---+-----l
o 0.1 0.2 0.3 0.4 0.5
Volume fraction of filler (-)
Figure 1.42 Effect of interaction on composite properties. Yield stress of PP-
CaC0 3 composites of Figure 1.9 presented in a linearized form.
Composition dependence of properties: Prediction 51
loadings around 25-30 vol%. The difference in elongation makes prediction
of strength difficult; at large elongations the cross-section of the specimen
changes and orientation of the matrix leads to strain hardening.
Modification of Equation 19 could successfully cope with these problems
and led to the following expression:
1-<PI
aT = aTm An 1 2 exp(B" <PI ) (21)
- .5 <PI
in which introduction of true tensile strength (aT =aA, A=L/Lo, relative
elongation) takes into account the change of specimen cross-section and An
accounts for strain hardening. n characterizes the strain hardening tendency
of the polymer and can be determined from matrix properties. Ba can be
related to interaction, similarly to Equation 20.
The correlation proved to be valid in all PP composites investigated to
date. Besides showing the effect of changing interaction, it sensitively
indicates changes in matrix properties and internal structure (aggregation)
or in the failure process. Figure 1.43 presents the composition dependence of
the tensile strength for three PP composites in linearized form. The filler
with 1.3)lm average particle size shows moderate reinforcing effect and an
intersection, which corresponds to the matrix value. In the case of large
particles (58 11m) interaction is weak and the dominant deformation mechan-
ism is debonding followed by intermediate failure. Small particles (0.08 )lm)
form aggregates, which initiate cracks. Although interaction is strong, failure

4.5

4.25
I

..c
C,
c
~
1ii
~ 3.75
.iii
c
!
""0
Q)
<J
:J
3.5
""0
~

.= 3.25

3
0 0.1 0.2 0.3 0.4 0.5
Volume fraction of filler ( - )
Figure 1.43 Effect of filler characteristics and structure on the strength of PP
composites. Particle size: (0) 1.3 Ilm, (0) 58 Ilm, (6) 0.01 Ilm.
52 Particulate filled polypropylene: structure and properties
takes place at low deformations. Changing slope (parameter B,,) and
intersection clearly indicate these phenomena. The model sensitively detects
heterogeneities and shows the effect of processing on structure and proper-
ties of composites.
Impact resistance and fracture are clearly major issues in both the
development and application of particulate filled composites. A large variety
of fracture properties are measured in PP composites and the most diverse
composition dependencies are observed. Although impact resistance usually
decreases with increasing amount of filler [205, 206J quite frequently a
maximum is observed at low or intermediate « 15 vol%) filler content [79,
156, 203, 206]. The maximum is most probably the combined result of
different competitive micromechanical deformation processes influenced by
polymer-filler and particle-particle interactions.
Because of the complexity of the phenomena taking place during the
fracture of particulate filled PP composites, the number of models predicting
composition dependence is extremely low. Kendall [47J developed a simple
correlation, which expresses composition dependence of the fracture resis-
tance of composites as the sum of matrix and interface toughnesses. The
same equation was used, in a slightly modified form, by Friedrich and
Karsch [114J for Si0 2 filled PP composites:
(22)
where Gc is the strain energy release rate of the composite, R~ and Rf the
crack resistance of the matrix and the interface, respectively, c the interfacial
contribution to resistance, and n' a geometric factor to take into account the
extra path of the crack around the particles. Kendall and Friedrich claim
that Rf~R~, thus the second term can be neglected and the equation takes
the form:
(23)
It is obvious, however, that this simple correlation is unable to predict the
widely differing composition dependencies, which range from exponential
decrease to correlations containing extremes (Figure 1.17). Further work is
needed on the development of appropriate models which take into account
the effect of different micromechanical deformation processes, the overall
plastic deformation of the matrix and the effect of interaction. It should be
noted that all the models break down when particle-particle interactions
reach a certain extent; the models are, however, quite helpful in detecting
them.

1.9 MULTICOMPONENT PP SYSTEMS


Introduction of a filler or reinforcement into a polymer improves some of its
properties, but has a detrimental effect on others. It is obvious that the
Multicomponent PP systems 53
properties of particulate filled composites can be improved by the combina-
tion of the advantageous properties of two or more different additives. PP is
no exception; numerous references can be found in the literature concerning
the application of more than one secondary component in a PP matrix.
These composites can be divided into two categories: hybrids containing
two different fillers or a filler and a reinforcement; and composites contain-
ing another polymer, usually elastomer, and a filler or reinforcement.
Hybrids are used to achieve synergistic effects and an improvement of
mechanical characteristics and functional properties or simply to decrease
price. Bertelli et al. [13, 207] studied the effect of particulate fillers on the
effectiveness of intumescent flame retardants and on the structure of the
insulating crust developed during burning. They found that fillers change
the structure of the crust, but influence disadvantageously the effectiveness
of the flame retardants. TormaHi, Paakkonen and Laiho [61] combined
wood flour and particulate fillers, talc and CaC0 3 , to improve mechanical
properties, dimensional stability and water adsorption of the composites.
They found improvement in the above properties, without any loss in
processing characteristics or economy.
Fiber reinforced PP composites have many advantageous properties and
in some areas they compete successfully even with engineering thermoplas-
tics. They have, however, two major drawbacks, which in some cases limit
their application: strong dependence of shrinkage on direction due to fiber
orientation, and price. To circumvent these problems occasionally particu-
late fillers are added to the fiber reinforced PP composite [5, 208]. The most
successful attempt to improve properties by this technique was reported by
Trotignon et al. [80]. They combined short glass fibers and mica and found
a synergetic effect of these additives on process ability, tensile strength,
impact resistance and heat deflection temperature, and the price of the
composites was also lower. A certain decrease of the modulus was observed
on the introduction of mica, but this was strongly outweighed by the
improvement of other properties and the economic advantage.
The number of publications on PP-polymer-filler composites is much
larger than those on hybrid systems. PP-PE-filler composites form one
family of these materials studied in some research groups. Arroyo and
colleagues [209-213] published a series of papers dealing with the melting
and crystallization phenomena, processing characteristics and mechanical
properties of such systems. Composition of PP and PE changes in the whole
composition range and filler content is varied at two levels. They also
studied the effect of processing technique (injection versus compression
molding) on the crystalline structure and properties of these composites.
They observed an improvement in the homogeneity and properties of the
composites prepared by injection molding over the compression molded
ones. Talc and glass fibers influenced the crystallization and crystalline
structure of PP, but not that of PE. Impact resistance decreased with
54 Particulate filled polypropylene: structure and properties
increasing PP content. Glass fiber attrition depended on composition and
processing technology and increased with glass fiber content. These obser-
vations are more or less in accordance with expectations. They did not
investigate, however, the distribution of the components and the dispersed
morphology of the composites, which have a crucial effect on properties.
Although they observed a small shift in phase transition with changing filler
characteristics and filler content, the distribution of the filler in the two
polymers is not clear. This question is not mentioned in the papers of Kerch
and Irgens [85J and Gupta and Gupta [214J either. Also, the goal and
expected advantages of the combination of PP, PE and the fillers remain
unclear.
Distribution of the filler in incompatible polymer blends, however, has a
large impact on properties, proved by the numerous papers on PP-
elastomer-filler systems and by the experiments of Sumita et ai. [63]. They
added carbon black to the incompatible blend of PP and PMMA. Two
different structures were observed; carbon black was found in quantities
larger than the average composition either in one of the polymers or at the
interface. Proper choice of composition and structure lead to enhanced
conductivity at lower carbon black content than in either of the components
alone.
The main class of multicomponent PP systems is formed by composites
containing PP, an elastomer and a filler at the same time. PP has a poor
low temperature impact strength, which is frequently improved by the
introduction of elastomers [96, 215]. Improvement in impact strength,
however, is accompanied by a simultaneous decrease of modulus, which
cannot be accepted in certain applications; a filler or reinforcement is
added to compensate for the effect. A large number of such systems were
prepared and investigated, but the observations concerning their structure,
the distribution of the components and their effect on composite properties
are rather controversial. In some cases separate distribution of the com-
ponents and independent effects were observed [216-218J, in others en-
capsulation of the filler by the elastomer [219-222]. Enrichment of the
interface in filler has not been mentioned in these publications. These
different structures naturally lead also to dissimilar properties. Figures 1.44
and 1.45 show the two structures mentioned above (separate dispersion,
encapsulation), while Figures 1.46 and 1.47 show the related differences in
properties. Composition dependence of shear modulus agrees well with
values predicted by the Lewis-Nielsen model in the case of separate
dispersion of the components (Figure 1.46), while large deviations are
observed when filler particles are embedded in the elastomer (Figure 1.47)
[223]. Embedding of the filler particles were also observed on the dynamic
mechanical spectra of the composites; encapsulated filler particles extended
the elastomer, leading to an increased relaxation peak maximum at around
-50 DC.
Multicomponent PP systems 55

Figure 1.44 Separate dispersion of the components in three-component PP-


EPDM-CaC0 3 composites. Composition: PP-EPDM-CaC0 3 , 60/20/20 volume
fraction.

Figure 1.45 Encapsulation of filler particles into the elastomers in three-compo-


nent PP-EPDM-CaC0 3 composites. Composition: PP-EPDM-CaC0 3 , 60/20/20
volume fraction.

Unfortunately, no direct methods are available which could give a


quantitative measure of the extent of encapsulation or the thickness of the
elastomer layer surrounding the filler particles. Comparison of measured
and calculated moduli, demonstrated by Figures 1.46 and 1.47, offers at
least a semi-quantitative estimate of the degree of encapsulation. The
56 Particulate filled polypropylene: structure and properties

2.-----~~-----.------T7--.__.

~ 1.5
(!)

(!)
u)
:::l
:;
"0
o
E
Q)
Ol
~
o : I
U5 0.5 - ............................
- i -····lI . ·······················_·····I·················...............
1".........................
!

i I !
! i I

o+-----~------~------+-----~
I I I
o 0.1 0.2 0.3 0.4
Volume fraction of filler ( - )
Figure 1.46 Composition dependence of storage modulus in PP-EPDM-CaC0 3
composites. Separate dispersion. Elastomer content: (0) 0, (.0,) 5, (D) 10, (\7) 20 vol%.
(- - - -) Lewis-Nielsen prediction.

m
Il..
1.5
(!)

(,
u)
:::l
~ .,. ........+...........................

t
+-'
en
~- !
0.5 f"'"~--t'......---i~;o;;;;;;;;;;;;;
- -- ... -
. . ;. ;. :·±·l··......·........·· . ·
.1.

o+-----~------~------+-----~
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)
Figure 1.47 Effect of filler encapsulation on the storage modulus of PP-EPDM-
CaC0 3 composites. Elastomer content: (0) 0, (.0,) 5, (D) 10, (\7) 20 vol%.
Muiticomponent PP systems 57
Lewis-Nielsen model extended to the three-component case describes the
composition dependence of the modulus at completely separate dispersion
of the components. When the filler is encapsulated its apparent amount in
the composite decreases and that of the elastomer increases. If the equation
is fitted to the measured modulus values an apparent volume fraction can be
calculated for both components. The difference between the effective and the
apparent volume fractions gives an estimate for the extent of encapsulation
[223,224].
Composition differences calculated according to the technique described
above are plotted in Figure 1.48 for the composite with the encapsulated
filler particles. An increase in both filler and elastomer content promotes
encapsulation. The most extensive encapsulation was observed at CPe = 0.2,
indicating that the relative amounts of the components play an important
role in the formation of this structure. These and similar calculations have
shown that in the investigated systems a maximum 70% of the filler
particles could be embedded, while even in the cases when separate
dispersion of the components dominates, at least 5-10% of the particles
were encapsulated. Attention must be drawn to the fact that the calculated
apparent volume fractions are relative values; they depend on the reference
equation used and are largely influenced by the values of parameters, as was
shown earlier.

0.3..--------,-------..,.------....,....------,

~
0.2 .................. ····r···
I
III
~ 0.1 ....................
~
tJ

:i
c:

0 t-E~~p;;~~~:::¥:;f~:::~
a.
E
o
tJ -0.1
III
>
.~

Qi
a:: -0.2

-0.3+-----~------~------+-----~
o 0.1 0.2 0.3 0.4
Volume fraction of filler (-)

Figure 1.48 Estimation of encapsulation. Relative composition differences cal-


culated with the extended Lewis-Nielsen model. Elastomer content: (0) 0, (6) 5,
(0) 10, (v) 20vol%.
58 Particulate filled polypropylene: structure and properties
The questions obviously arise as to what is the reason for the different
morphologies and which conditions favor one or the other. The role of
collision probability, energetics and stability of the formed structure were
studied by Puk:inszky, Kolatik and colleagues [166, 218, 223-227] to
answer these questions and to analyze structure formation. They have
shown that collision probability has a secondary role, proven also by the
independence of the structure on the sequence of composite preparation.
Previous homogenization of any pair of components and introduction of the
third lead to the same structure [218].
Model calculations were carried out in order to check the role of
energetics in structure formation. To create new surfaces surplus energy is
needed and the structure which requires less energy will form predominant-
ly. Using some simplifying assumptions the surplus energy necessary to form
the new surfaces in the extreme cases, i.e. 100% embedding and 100%
separate dispersion, was calculated. The results, which are listed in Table 1.7,
show that in both cases - surface treated and non-treated filler - encapsula-
tion is the thermodynamically favored process [166, 223].

Table 1.7 Surplus offree energy necessary to create new surfaces in a unit volume of
PP-EPDM-CaC0 3 composites having encapsulated and separately dispersed mor-
phology. Composition: PP-EPDM-CaC0 3 60/20/20 vol %

Surplus free energy (kJ/m3)

Filler Separately dispersed Encapsulated


treatment L.\Us L.\U E

Non-treated 236.5 203.9 32.6


Treated 24.5 15.0 9.5

The final structure of the composites is formed during processing. The


stability of the multilayer, embedded units is determined by the adhesion
between the layers, i.e. filler and elastomer, and by the shear forces
developing in the melt. Although Equation 13 was originally developed in
order to describe interaction between two particles or a particle and a plate
[164, 165], it was adapted to the studied case, and adhesion between the
elastomer layer and the filler particle was calculated. Shear forces were
estimated by the technique of Goodrich and Porter [167] and they were
shown to be in the same order of magnitude as adhesion. The results of the
model calculations are plotted in Figure 1.49. With surface treatment,
interfacial tension and work of adhesion decrease (see Figure 1.25), while
shear stress changes only slightly with composition. Decrease of adhesion
can reach an extent where it is surpassed by shear forces; de-encapsulation
of the particles takes place. Equation 13 shows, however, that particle size is
an equally important factor. Fillers usually have a wide particle size
Muiticomponent PP systems 59

Total additive content (-)


o 0.1 0.2 0.3 0.4 O.S
SO.---~-----+----~--~~---.SO

45 .... 45

-;
i::E 40 0.
e
-c
0
'i
35
III
III
t'!
';
ell
.t::. ...as
'C
« 30
ell
30 .t::.
(J)

,
2S ······_··············f 25

20 20
0 20 40 60 80 100
Surface coverage ('10)
Figure 1.49 Change of adhesive forces and shear stress with surface treatment and
composition, respectively. Role of their relative magnitude in structure formation.
(0) EPDM-CaC0 3, (0) PP-CaC0 3 adhesion, (6) shear stress.

distribution, very often in the 0.3-10 Ilm range. As a consequence, compos-


ites will always contain small particles that are encapsulated and large
particles that are de-encapsulated, despite the presence or absence of surface
treatment [166].
The analysis of structure formation and Equation 13 show that the
determining factor is the relative magnitude of adhesion and shear forces,
which is determined by interaction of the components, particle size and
shear stress during homogenization. Experiments to investigate the effect
of these factors have shown that increasing surface coverage of filler with
a surfactant, i.e. decreasing reversible work of adhesion, invariably leads
to decreasing encapsulation (Figure 1.50). Similarly, decreasing particle
size increases the amount of embedded particles (Figure 1.51). In the case
of large particles (58J.1m) hardly any encapsulation occurs, while very
small particles cannot be de-encapsulated even at high (100%) surface
coverage [225]. Also, the effect of shear stress could be qualitatively
proved in similar model experiments [225]. These results are in accord-
ance with the tentative explanation presented above, which is further
corroborated by the observation of Stamhuis [219-221] and Varga [42].
Stamhuis [219-221] has shown that in some cases only partial encapsu-
lation occurs, and the degree of encapsulation depends on the polarity of
the elastomer, i.e. on adhesion. Varga [42] has made a similar observation
60 Particulate filled polypropylene: structure and properties
1.2 . , - - - - - , - - - - y - - - - . - - - - - ,

'i
~
(!j
ui
:::l
:; 0.8
-g
E
CD
C)
~
g
CJ) 0.6

0.4 +-__-+_ _ _+ __--t_ _ _...J


o 50 100 150 200
Surface coverage (%)
Figure 1.50 Effect of filler surface treatment (interaction, Wmf) on the shear
modulus of PP-EPDM-CaC0 3 composites. Increasing de-encapsulation with in-
creasing surface coverage. Composition PP-EPDM-CaC0 3: (0) 50/20/30, (c.)
60/20/20, (0) 70/20/lOvol%.

100

~ 80
(!j

~.""-
<I
a)
I.)
c:
60
I!! v
~
_...._----
~-----. -----
'6 40
Ul
:::l
:;
"C
0
E 20
CD

~!~
.~
co
"ai
a: 0

-20
o 50 100 150 200
Surface coverage (%)
Figure 1.51 Relative drop in the modulus as a function of surface treatment
normalized to complete surface coverage. Particle size: (0) 0.081JlIl, (c.) 3.61JlIl,
(0) 58 1JlIl.
Appendix 61
concerning encapsulation; moreover, he has shown that even a very small
amount of filler (1 wt%) cannot be covered completely with as much as
10 wt% elastomer.
All of these experiments have shown the primary importance of ad-
hesion in structure formation and in the resulting effect on properties.
Numerous attempts have been made to prepare composites with exclusive
structure, i.e. complete coverage or separate distribution. In most cases PP
or elastomer modified with maleic anhydride or acryclic acid was used to
enhance adhesion between selected components; separate dispersion in the
case of MA-PP, embedding with MA-EPDM [159, 160, 227, 228]. Al-
though the desired effect was observed in almost every case, the exclusive-
ness could not be proved. Similarly contradictory is the effect on proper-
ties, especially on impact resistance. In three-component PE composites
better impact resistance/modulus ratio could be achieved, but Kelnar
[159] has observed lower impact strength with acrylic-acid modified
EPDM (AA-EPDM) than with a non-modified elastomer. Scott and
Ishida have also emphasized the complexity of the problem and the
necessity of further study to develop composites with optimum properties
[229].

1.10 APPENDIX:
SYMBOLS AND ABBREVIA nONS

c Interfacial contribution to resistance against crack propagation


Maximum amount of surfactant bonded on the filler surface
Amount of irreversibly bonded surfactant
Inverse gas chromatography
Einstein's coefficient
Thickness of the interphase
n Strain hardening tendency of the matrix
n' Geometry factor for extra path in crack propagation
u Deformation vector
A Parameter related to filler anisotropy
AA Acrylic-acid
Ar Specific surface area of the filler
B Parameter of the Lewis-Nielsen equation [71]
Buy Correlation parameter calculated from yield stress
Bu Correlation parameter calculated from tensile strength
C l , C2 Constants
E Young's modulus
ESCA Electron spectroscopy for chemical applications
EPDM Ethylene-propylene-diene terpolymer
ESR Electron spin resonance
62 Particulate filled polypropylene: structure and properties
G Storage shear modulus
Gc Critical strain energy release rate
GPC Gel permeation chromatography
L1Hc Heat of crystallization
IGC Inverse gas chromatography
K1 Constant
PE Polyethylene
PP Polypropylene
MA Maleic-anhydride
R Particle radius
RC Crack resistance
Rm Ratio of component shear moduli, Gm/Gr
SALS Small-angle light scattering
SEM Scanning electron microscopy
Smr Wettability
L1S New surface created in debonding
Tc Crystallization peak temperature
L1T Temperature difference giving rise to thermal stresses
U Elastic energy
I1U Surplus of free energy necessary to create new surfaces III
multicomponent PP composites
Reversible work of adhesion
Thermal expansion coefficient
Surface tension
Dispersion coefficient of surface tension
Polar component of surface tension
Interfacial tension
Yield strain
Relative elongation
v Poisson's ratio
Pr Density of the filler
a Stress tensor
0"3 Specific adhesion force
O"D Debonding stress
O"y Tensile yield stress
0"1,0"2,0"3 Principal stress components
O"T Thermal stress
CPr Volume fraction of filler
cp}"3X Maximum packing fraction
'P Parameter related to cp}"3X

Superscripts and subscripts m, f and i stand for matrix (polymer), filler and
interface or interphase.
References 63
1.11 REFERENCES
1. Schlumpf, H.-P. (1990) Modern aspects of fillers in polypropylene, presented at
the 10th International Macromolecule Symposium, 20-21 September, Inter-
laken, Switzerland.
2. Vink, D. (1990) KunststoJfe, 80, 842-6.
3. Aumayr, G. (1989) Kunststojjberater, 34, 63-6.
4. Schlumpf, H.-P. and Bilogan, W. (1981) KunststoJfe, 71, 820-3.
5. Schlumpf, H.-P. (1983) KunststoJfe, 73, 511-15.
6. Schmidt, H. and Jzquierdo, R. (1988) Kunststoff, 78, 149-50.
7. Malpass, V. and Kempthorn, J. (1989) Plastics Compounding, 12, 52, 55-6,
58-61.
8. Bezard, D. (1986) Oesterreichische KunststoJf Zeitschrift, 17, 174.
9. Aumayr, G. and Neifjl, W. (1987) Oesterreichische KunststoJf Zeitschrift, 18,
121-2, 124.
10. Stadterman, R. L. and Willis, A. 1. (1988) TAPP] Journal, 71,145-6.
11. Bihlmayer, G. A., Aumayr, G. and Mautner, F. (1988) Plaste und Kautschuk, 35,
185-9.
12. Domininghaus, H. (1984) Gummi, Asbest, KunststoJfe, 37, 326-30.
13. Bertelli, G., Camino, G., Marchetti, E. et al. (1989) Polymer Degradation and
Stability, 25, 277-92.
14. Acosta, J. L., Rodriguez, M., Linares, A. and Jurado, J. R. (1990) Polymer
Bulletin, 24, 87-91.
15. Bricteux, 1. and Michaux, C. (1971) Proprietes et caracteristiques des charges
inorganiques, Bulletin Technique 'Polymeres', No.1, Institut National des
Industries Extractives, Liege, Belgium.
16. Samuels, R. 1. (1974) Structured Polymer Properties, Wiley, New York.
17. Greco, R. and Coppola, F. (1986) Plastics and Rubber Processing and Applica-
tions, 6, 35-41.
18. Wright, D. G. M., Dunk, R., Bouvart, D. and Autran, M. (1988) Polymer, 29,
793-8.
19. Trotignon, J. P. and Verdu, J. (1987) Journal of Applied Polymer Science, 34,
1-18.
20. Fujiyama, M., Wakino, T. and Kawasaki, Y. (1988) Journal of Applied Polymer
Science, 35, 29-49.
21. Fujiyama, M. and Wakino, T. (1991) Journal of Applied Polymer Science, 42,
2749-60.
22. Singh, P. and Kamal, M. R. (1989) Polymer Composites, 10, 344-51.
23. Menges, G., Troost, A., Koske, 1. et al. (1988) KunststoJfe, 78, 806-9.
24. Karger-Kocsis, J. and Friedrich, K. (1989) International Journal of Fatigue, 11,
161-8.
25. Murphy, M. W., Thomas, K. and Bevis, M. J. (1987) Plastics and Rubber
Processing and Applications, 7, 241-2.
26. Murphy, M. W., Thomas, K. and Bevis, M. J. (1988) Plastics and Rubber
Processing and Applications, 9, 3-16.
27. Trotignon, J. P. and Verdu, J. (1987) Journal of Applied Polymer Science, 34,
19-36.
28. Bartosiewicz, L. and Kelly, C. 1. (1987) Advances in Polymer Technology, 7,
21-33.
29. Riley, A. M., Paynter, C. D., McGenity, P. M. and Adams, J. M. (1990) Plastic
and Rubber Processing and Applications, 14, 85-93.
30. Menczel,1. and Varga, 1. (1983) Journal of Thermal Analysis, 28,161-74.
64 Particulate filled polypropylene: structure and properties
31. Fujiyama, M. and Wakino, T. (1991) Journal of Applied Polymer Science, 42,
2739-47.
32. Varga, J. Kemiai Kozlemenyek, 71, 285-309 (Hungarian).
33. Varga, J. and Schulek-T6th, F. (1991) Angewandte Makromolekulare Chemie,
188, 11-25.
34. Varga, J. (1989) Journal of Thermal Analysis, 35, 1891-1912.
35. Kowalewski, T. and Galeski, A. (1986) Journal of Applied Polymer Science, 32,
2919-34.
36. Ribnikcit, F. (1991) Journal of Applied Polymer Science, 42, 2727-37.
37. Garton, A., Kim, S. W. and Wiles, D. M. (1982) Journal of Polymer Science,
Polymer Letters, 20, 273-8.
38. Zuchowska, D. and Hlavata D. (1991) European Polymer Journal, 27, 355-7.
39. Bajaj, P., Jha, N. K., Maurya, P. L. and Misra, A. C. (1987) Journal of Applied
Polymer Science, 34, 1785-1801.
40. Yue, C. Y. and Cheung, W. L. (1991) Journal of Materials Science, 26, 870-80.
41. Xavier, S. F. and Sharma, Y. N. (1984) Angewandte Makromolekulare Chemie,
127, 145-52.
42. Varga, J. (1991) Journal of Polymer Engineering, 1,231-51.
43. Folkes, M. 1. and Wong, W. K. (1987) Polymer, 28, 1309-14.
44. Folkes, M. 1. and Hardwick, S. T. (1987) Journal of Materials Science Letters, 6,
656-8.
45. Burton, R. H., Day, T. M. and Folkes, M. J. (1984) Polymer Communications,
25,361-2.
46. Huson, M. G. and McGill, W. J. (1984) Journal of Polymer Science, Polymer
Chemistry, 22, 3571-80.
47. Kendall, K. (1978) British Polymer Journal, 10, 35-8.
48. Maiti, S. N. and Mahapatro, P. K. (1989) Journal of Applied Polymer Science,
37,1889-9.
49. Maiti, S. N. and Mahapatro, P. K. (1990) International Journal of Polymeric
Materials, 14, 205-222.
50. Hutley, T. J. and Darlington, M. W. (1984) Polymer Communications, 25, 226-8.
51. Hutley, T. 1. and Darlington, M. W. (1985) Polymer Communications, 26, 264-7.
52. Schlumpf, H.-P. (1990) Chimia, 44, 359-60.
53. Golovoy, A. (1986) Polymer Composites, 7, 405-12.
54. Voelker, M. J. (1991) Polymer Composites, 12, 119-21.
55. Vollenberg, P. H. T. and Heikens, D. (1986) The effects of particle size on the
mechanical properties of composites, in Composite Interfaces, (eds H. Ishida
and 1. L. Koenig), Elsevier, New York, pp. 171-5.
56. Ramsteiner, F. and Theysohn, R. (1984) Composites, 15, 121-8.
57. Morales, E. and White, 1. R. (1988) Journal of Materials Science, 23, 3612-
22.
58. Arroyo, M., Iglesias, A. and Perez, F. (1985) Journal of Applied Polymer Science,
30, 2475-83.
59. Miyata, S., Imahashi, T. and Aabuki, H. (1980) Journal of Applied Polymer
Science, 25, 415-25.
60. Hornsby, P. R. and Watson, C. L. (1989) Plastic and Rubber Processing and
Applications, 11,45-51.
61. Tormiilii, P., Piiiikkonen, E. and Laiho, 1. (1985) Kunststoffe, 75, 287-90.
62. Socha, D. A. (1989) Calcium sulfate - the performance additive, in Filplas
Conference, 12-13 April, Manchester, UK.
63. Sumita, M., Sakata, K., Asai, S. et al. (1991) Polymer Bulletin, 25, 265-71.
64. Bigg, D. M. (1987) Polymer Composites, 8, 115-22.
References 65
65. Bridge, B., Folkes, M. 1. and Jahankhani, H. (1988) Journal of Materials
Science, 23, 1948-54.
66. Nagy, T. T., Vamos, Gy., T6th, A. et al. (1984) Miianyag es Gumi, 21, 48-52
(Hungarian).
67. Ishida, H. and Miller, J. D. (1984) Macromolecules, 17, 1659-66.
68. Chen, L.-S., Mai, Y.-W. and Cotterell, B. (1989) Polymer Engineering and
Science, 29, 505-12.
69. Mitsuishi, K., Kodama, S. and Kawasaki, H. (1985) Polymer Engineering and
Science, 25, 1069-73.
70. Michler, G. H. and Tovmasjan, 1. M. (1988) Plaste und Kautschuk, 35, 73-7.
71. Nielsen, L. E. (1974) Mechanical Properties of Polymers and Composites, Marcel
Dekker, New York.
72. Suetsugu, Y., Kikutani, T., Kyu, T. and White, 1. L. (1990) Colloid and Polymer
Science, 268, 118-31.
73. Suetsugu, Y. and White, J. L. (1987) Advances in Polymer Technology, 7, 427--49.
74. Svehlova, V. and Poloucek, E. (1987) Angewandte Makromolekulare Chemie,
153, 197-200.
75. Pukanszky, B., Turcsanyi, B. and Tiid6s, F. (1988) Effect of interfacial interac-
tion.on the tensile yield stress of polymer composites, in Interfaces in Polymer,
Ceramic, and Metal Matrix Composites (ed. H. Ishida), Elsevier, New York,
pp.467-77.
76. Pukanszky, B. (1990) Composites, 21, 255-62.
77. Busigin, c., Lahtinen, R., Martinez, G. M. et al. (1984) Polymer Engineering and
Science, 24, 169-74.
78. Parrinello, L. M. (1991) TAPPI Journal, 74, 85-8.
79. Vu-Khanh, T. and Fisa, B. (1986) Polymer Composites, 7, 219-26.
80. Trotignon, 1. P., Verdu, J., De Boissard, R. and De Vallois, A. (1986)
Polypropylene-mica composites, in Polymer Composites (ed. B. Sedlatek),
Walter de Gruyter, Berlin, pp. 191-8.
81. Pukanszky, B., Fekete, E. and Tiid6s, F. (1989) Makromolekulare Chemie,
Macromolecular Symposia, 28, 165-86.
82. Ok uno, K. and Woodhams, R. T. (1975) Polymer Engineering and Science, 15,
308-15.
83. Pukanszky, B., Tiid6s, F., Jantat, J. and Kolatik, J. (1989) Journal of Materials
Science Letters, 8, 1040--42.
84. Stoklasa, K., Tomis, F. and Navratil, Z. (1985) Thermochimica Acta, 93, 221--4.
85. Kerch, G. M. and Irgens, L. (1985) Thermochimica Acta, 93, 155-7.
86. Bridge, B., Folkes, M. J. and Jahankhani, H. (1989) Journal of Materials
Science, 24, 1479-85.
87. Goodier, J. N. (1933) Journal of Applied Mechanics, 55, 39--44.
88. Kowalewski, T., Galeski, A. and Kryszewski, M. (1984) The structure and
tensile properties of cold drawn modified chalk filled polypropylene, in Polymer
Blends. Processing, Morphology and Properties (eds M. Kryszewski, A. Galeski
and E. Martuscelli), Plenum Press, New York, pp. 223--41.
89. Nakagawa, H. and Sano, H. (1985) Polymer Preprints, 26, 249-50.
90. Maiti, S. N. and Mahapatro, P. K. (1991) Journal of Applied Polymer Science,
42, 3101-10.
91. Trantina, G. G. (1984) Polymer Engineering and Science, 24, 1180-84.
92. Nicolais, L. and Nicodemo, L. (1974) International Journal of Polymeric
Materials, 4, 229--43.
93. Vollenberg, P. H. T. (1987) The mechanical behaviour of particle filled
thermoplastics, PhD Thesis, Eindhoven University of Technology.
66 Particulate filled polypropylene: structure and properties
94. van der Sanden, M. C. M. and Heikens, D. (1990) Influence of adhesion on
shear band formation around LDPE particles in polycarbonate, presented at
the 13th Discussion Conference on the Mechanisms of Polymer Strength and
Toughness, 16-19 July, Prague, Czechoslovakia.
95. Puklinszky, B. and Voros, Gy. (1993) Composite Interface, 1,411-27.
96. Bucknall, C. B. (1977) Toughened Plastics, Applied Science, London.
97. Nezbedova, E. and Davidovic, I. (1986) Plaste und Kautschuk, 33, 228-30.
98. Beck, R H., Gratch, S., Newman, S. and Rausch, K. C. (1968) Journal of
Polymer Science, Polymer Letters, 6, 707-9.
99. Galeski, A. and Kalinski, R (1980) Polymeric modifier for filled polypropylene,
in Polymer Blends. Processing, Morphology and Properties Vol. 1, (eds. E.
Martuscelli, R. Palumbo and M. Kryszewski) Plenum Press, New York,
pp.431-49.
100. Wu, S. (1974) Journal of Macromolecular Science, Reviews in Macromolecular
Chemistry, CI0, 1-73.
101. Dekkers, M. E. J. and Heikens, D. (1986) Interfacial effects on local deforma-
tion mechanisms in glass bead-filled glassy polymers, in Composite Interfaces
(eds H. Ishida and 1. L. Koenig), Elsevier, New York, pp. 161-70.
102. Dekkers, M. E. J. and Heikens, D. (1985) Journal of Materials Science, 20,
3873-80.
103. Jang, B. Z., Uhlmann, D. R and Van der Sande, 1. B. (1985) Polymer
Engineering and Science, 25, 98-104.
104. Kinloch, A. 1. and Young, R. 1. (1983) Fracture Behaviour of Polymers, Elsevier,
London.
105. Breuer, H. (1985) Yielding and failure criteria for rubber modified polymers,
Part 1, NATO ASI Series, Series E, 89, 375-82.
106. Breuer, H. (1985) Yielding and failure criteria for rubber modified polymers,
Part 2, NATO ASI Series, Series E, 89, 383-92.
107. Dekkers, M. E. J. and Heikens, D. (1984) Journal of Materials Science, 19,
3271-275.
108. Vollenberg, P., Heikens, D. and Ladan, H. C. B. (1988) Polymer Composites, 9,
382-8.
109. Puklinszky, B., van Es, M., Maurer, F. H. J. and Voros, Gy. (1994) Journal of
Materials Science, 29, 2350-8.
110. Allard, R. c., Vu-Khanh, T. and Chalifoux, J. P. (1989) Polymer Composites, 10,
62-8.
111. Vu-Khanh, T. and Fisa, B. (1986) Polymer Composites, 7, 375-82.
112. Soya, M. (1986) Strain rate dependence of deformation behaviour and fracture
surface morphology in polypropylene with short glass fibres, in Polymer
Composites (ed. B. Sedlacek), Walter de Gruyter, Berlin, pp. 507-14.
113. Vu-Khanh, T., Sanschagrin, B. and Fisa, B. (1985) Polymer Composites, 6,
249-60.
114. Friedrich, K. and Karsch, U. A. (1983) Fibre Science and Technology, 18,
37-52.
115. Hoffmann, H., Grellmann, W. and Zilvar, V. (1986) Instrumental impact studies
of some thermoplastic composites, in Polymer Composites (ed. B. SedIM:ek),
WaIter de Gruyter, Berlin, pp. 233-42.
116. Nezbedova, E., Ponesicky, J. and So va, M. (1990) Acta Polymerica, 41, 36-42.
117. Vollenberg, P. H. T. and Heikens, D. (1989) Polymer, 30, 1656-62.
118. Maurer, F. H. J., Kosfeld, R, Uhlenbroich, T. and Bosveliev, L. G. (1981)
Structure and properties of highly filled high-density polyethylene, in 27th
International Symposium on Macromolecules, 6-9 July, Strasbourg, France.
References 67
119. Maurer, F. H. J., Schoffeleers, H. M., Kosfeld, R. and Uhlenbroich, T. (1982)
Analysis of polymer-filler interaction in filled polyethylene, in Progress in
Science and Engineering of Composites, ICCM-IV (eds T. Hayashi, K. Kawata
and S. Umekawa), Tokyo, pp. 803-9.
120. Akay, G. (1990) Polymer Engineering Science, 30, 1361-72.
121. Maurer, F. H. 1., Kosfeld, R. and Uhlenbroich, Th. (1985) Colloid and Polymer
Science, 263, 624--30.
122. Mansfield, K. F. and Theodorou, D. N. (1991) Macromolecules, 24, 4295-309.
123. Jancat, 1. (1991) Journal of Materials Science, 26, 4123-9.
124. Zorll, U. (1977) Gummi, Asbest, KunststofJe, 30, 436-44.
125. Iisaka, K. and Shibayama, K. (1978) Journal of Applied Polymer Science, 22,
3135-43.
126. Pukanszky, B. and Tiid6s, F. (1990) Indirect determination of interphase
thickness from the mechanical properties of particulate filled polymers, in
Controlled Interphases in Composite Materials (ed. H. Ishida), Elsevier, New
York, pp. 691-700.
127. Kolarik, J., Hudecek, S. and Lednicky F. (1978) Faserforschung und Textil-
technck., 29, 51-6.
128. Kolarik, J., Hudecek, S., Lednicky, F. and Nicolais, L. (1979) Journal of Applied
Polymer Science, 23, 1553-64.
129. Gupta, V. B., Mittal, R. K., Sharma, P. K. et al. (1989) Polymer Composites, 10,
16--27.
130. Sumita, M., Tsukini, H., Miyasaka, K. and Ishikawa, K. (1984) Journal of
Applied Polymer Science, 29, 1523-30.
131. Turcsanyi, B., Pukanszky, B. and Tiid6s, F. (1988) Journal of Materials Science
Letters, 7, 160--2.
132. Pukanszky, B. (1992) New Polymer Materials, 3, 205-17.
133. Felix, J. M. and Gatenholm, P. (1991) Journal of Applied Polymer Science, 42,
609-20.
134. Nicolais, L. and Nicodemo, L. (1973) Polymer Engineering and Science, 13,
469.
135. Schreiber, H. P., Wertheimer, M. R. and Lambla, M. (1982) Journal of Applied
Polymer Science, 27, 2269-80.
136. Fowkes, F. M. (1964) Industrial and Engineering Chemistry, 56, 40--52.
137. Fekete, E., Pukanszky, B., T6th, A. and Bert6ti, I. (1990) Journal of Colloid and
Interface Science, 135, 200--8.
138. Gutowski, W. (1990) Effect of fibre-matrix adhesion on mechanical properties
of composites, in Controlled Interphases in Composite Materials (ed. H. Ishida),
Elsevier, New York, pp. 505-20.
139. Huntsberger, J. R. (1981) Journal of Adhesion, 12,3-12.
140. Fox, H. W., Hare, E. F. and Zismann, W. A. (1955) Journal of Physical
Chemistry, 59, 1097-1106.
141. Matienzo, L. J. and Shah, T. K. (1980) Surface and Interface Analysis, 8, 53-9.
142. Raj, R. G., Kokta, B. V. and Daneault, C. (1989) International Journal of
Polymer Materials, 12, 239-50.
143. Monte, S. J. and Sugerman, G. (1990) Journal of Elastomers and Plastics, 22,
83-96.
144. Jancat, J. and Kucera, J. (1990) Polymer Engineering and Science, 30, 707-13.
145. Marosi, Gy., Bertalan, Gy., Rusznak, I. and Anna, P. (1986) Colloids and
Surfaces, 23, 185-98.
146. Papirer, E., Schultz, J. and Turchi, C. (1984) European Polymer Journal, 12,
1155-8.
68 Particulate filled polypropylene: structure and properties
147. Raj, R. G., Kokta, B. V., Dembe1e, F. and Sanschagrin, B. (1989) Journal of
Applied Polymer Science, 38, 1987-96.
148. Vollenberg, P. H. T. and Heikens, D. (1990) Journal of Materials Science, 25,
3089-95.
149. Mader, E. and Freitag, K.-H. (1990) Composites, 21, 397-402.
150. Tausz, S. E. and Chaffey, C. E. (1982) Journal of Applied Polymer Science, 27,
4493-500.
151. Plueddemann, E. P. (1982) Silane Coupling Agents, Plenum Press, New York.
152. Ishida, H. (1985) Structural gradient in the silane coupling agent layers and its
influence on the mechanical and physical properties of composites, in Molecu-
lar Characterization of Composite Interfaces (eds H. Ishida and G. Kumar),
Plenum Press, New York.
153. Favis, B. D., Blanchard, L. P., Leonard, 1. and Prud'homme, R. E. (1983)
Journal of Applied Polymer Science, 28, 1235-44.
154. Marshall, C. J., Rozett, R. and Kunkle, A. C. (1985) Plastics Compounding, 8,
69-70, 73-4.
155. Wright, P. 1. (1986) Developments in Plastics Technology, 3, 119-54.
156. Bajaj, P., Jha, N. K. and Jha, R. K. (1989) Polymer Engineering and Science, 29,
557-63.
157. Jancat, 1., Kummer, M. and Kolatik, J. (1988) Interfacial adhesion and
mechanical properties of maleated polypropylene/CaC0 3 composites, in Inter-
faces in Polymer, Ceramic, and Metal Matrix Composites (ed. H. Ishida),
Elsevier, New York, pp. 705-11.
158. Felix, J. M. and Gatenholm, P. (1991) Journal of Applied Polymer Science, 50,
699-708.
159. Kelnar, I. (1991) Angewandte Makromolekulare Chemie, 189,207-18.
160. Chiang, W.-Y. and Yang, W.-D. (1988) Journal of Applied Polymer Science, 35,
807-23.
161. Jancat, J. and Kucera, J. (1990) Polymer Engineering and Science, 30, 714-20.
162. Klason, C. and Kubat, J. (1986) Plastics and Rubber Processing and Applica-
tions, 6, 17-20.
163. Ess, J. W. and Hornsby, P. R. Plastics and Rubber Processing and Applications,
8,147-56.
164. Tabor, D. (1977) Journal of Colloid and Interface Science, 58, 2-13.
165. Roberts, A. D. (1979) Rubber Chemistry and Technology, 52, 23-42.
166. Pukanszky, B., Tiidos, F., Kolarik, J. and Lednicky, F. (1990) Polymer Compos-
ites, 11, 98-104.
167. Goodrich, J. E. and Porter, R. S. (1967) Polymer Engineering and Science, 7,
45-51.
168. Ess, J. W. and Hornsby, P. R. (1986) Polymer Testing, 6, 205-18.
169. Krapez, 1. c., Cie1o, P., Maldague, X. and Utracki, L. A. (1987) Polymer
Composites, 8, 396-407.
170. Bigg, D. M. (1983) Polymer Engineering and Science, 23, 206-10.
171. Kubat, J. and Szalanczi, A. (1974) Polymer Engineering and Science, 14, 873.
172. Utracki, L. A. (1984) The shear and e1ongational flow of polymer melts
containing anisometric filler particles; Part II, Mica-filled polyolefins, in
Proceedings ofthe IX International Congress on Rheology, Mexico, pp. 467-74.
173. Gupta, V. B., Mittal, R. K. and Sharma, P. K. (1989) Polymer Composites, 10,
8-15.
174. Kamal, M. R., Song, L. and Singh, P. (1986) Polymer Composites, 7, 323-9.
175. Sanou, M., Chung, B. and Cohen, C. (1985) Polymer Engineering and Science,
25, 1008-16.
References 69
176. Lisy, F., Hiltner, A. and Baer, E. (1991) Polymer Preprints, 32, 21-2.
177. Rockenbauer, A., J6kay, L., Pukimszky, B. and Tiidos, F. (1985) Macro-
molecules, 18, 918-23.
178. Fujiyama, M. and Wakino, T. (1991) Journal of Applied Polymer Science, 42,
9-20.
179. Mittal, R. K., Gupta, V. B. and Sharma, P. (1987) Journal of Materials Science,
22, 1949-55.
180. Christie, M. (1986) Plastics Engineering, 42, 41-5.
181. Sanschagrin, B., Gauvin, R., Fisa, B. and Vu-Khanh, T. (1987) Plastics
Compendivar, 10, 37-38,40, 42-3, 45, 48.
182. Fisa, K. B., Dufour, 1. and Vu-Khanh, T. Polymer Composites, 8, 408-17.
183. Vaxman, A. and Narkis, M. (1991) Polymer Composites, 12, 161-8.
184. Han, C. D. (1974) Journal of Applied Polymer Science, 18, 821-9.
185. Faulkner, D. L. and Schmidt, L. R. (1977) Polymer Engineering and Science, 17,
657-65.
186. Stamhuis, J. E. and Loppe, J. P. A. (1982) Rheologica Acta, 21, 103-5.
187. Jeffrey, D. J. and Acrivos, A. (1976) AIChE Journal, 22, 417-32.
188. Juskey, V. P. and Chaffey, C. E. (1982) Canadian Journal of Chemical Engineer-
ing, 60, 334--41.
189. Pukimszky, B., Tiidos, F. and Kelen, T. (1986) Polymer Composites, 7, 106--15.
190. Pukimszky, B. and Tiidos, F. (1988) Estimation of the interfacially bonded
polymer in PP composites, in Polymers for Advanced Technologies (ed.
M. Lewin), VCH Publishers, New York, pp. 792-807.
191. Utracki, L. A. (1984) Rubber Chemistry and Technology, 57, 507-22.
192. Kamal, M. R., Mutel, A. T. and Utracki, L. A. (1984) Polymer Composites, 5,
289-98.
193. McGee, S. and McCullough, R. L. (1981) Polymer Composites, 2, 149-61.
194. Dickie, R. A. (1978) Mechanical properties (small deformations) of multiphase
polymer blends, in Polymer Blends, Vol. 1, (eds D. R. Paul and S. Newman),
Academic Press, New York, pp. 353-91.
195. Ahmed, S. and Jones, F. R. (1990) Journal of Materials Science, 25,4933--42.
196. Christensen, R. M. and Lo, K. H. (1979) Journal of Mechanics and Physics of
Solids, 27, 315-30.
197. Lewis, T. R. and Nielsen, L. E. (1970) Journal of Applied Polymer Science, 14,
1449-71.
198. Mitsuishi, K., Kawasaki, H. and Kodama, S. (1984) Kobunshi Ronbunshu, 41,
665-72.
199. Katz, H. S. and Milewski, J. V. (1978) Handbook of Fillers and Reinforcements
for Plastics, Van Nostrand, New York.
200. Chacko, V. P., Karasz, F. E. and Farris, R. J. (1982) Polymer Engineering and
Science, 15,968-74.
201. Nicolais, L. and Narkis, M. (1971) Polymer Engineering and Science, 11, 194-9.
202. Ishai, O. and Cohen, L. J. (1968) Journal of Composite Materials, 2, 302-15.
203. Kokta, B. V., Raj, R. G. and Daneault, C. (1989) Polymer-Plastic Technology
and Engineering, 28, 247-59.
204. Nicolais, L. and Nicodemo, L. (1973) Polymer Engineering and Science, 13,469.
205. Bigg, D. M. (1987) Polymer Composites, 8, 115-122.
206. Kucera, J. (1986) Mechanical properties of polyethylene and polypropylene
filled with calcium carbonate, in Polymer Composites (ed. B. Sedhicek), Walter
de Gruyter, Berlin, pp. 544-52.
207. Bertelli, G., Marchetti, E., Camino, G. et al. (1989) Angewandte Mak-
romolekulare Chemie, 172, 153-63.
70 Particulate filled polypropylene: structure and properties
208. Acosta, J. L., Morales, E., Ojeda, M. C. and Linares, A. (1986) Angewandte
M akromolekulare Chemie, 138, 103-10.
209. Arroyo, M. and Vigo Matheu, J. P. (1988) Polymer Composites, 9, 105-11.
210. Arroyo, M., Sanchez-Berna, M. and Vigo Matheu, J. P. (1990) Journal of
Polymer Engineering, 9, 85-104.
211. Arroyo, M. and Avalos, F. (1989) Polymer Composites, 10, 117-21.
212. Arroyo, M. and Avalos, F. (1991) Polymer Composites, 12, 1-6.
213. Arroyo, M. and Avalos, F. (1991) Polymer Composites, 12, 7-1l.
214. Gupta, A. K. and Gupta, V. B. (1981) Polymer Composites, 2, 149-6l.
215. Karger-Kocsis, J., Kalla, A. and Kuleznev, V. N. (1984) Polymer, 25, 279-86.
216. Lee, Y.-D. and Lu, c.-c. (1982) Journal of the Chinese Institute of Chemical
Engineering, 13, 1-8.
217. Dao, K. C. and Hatem, R. A. (1984) SPE Technical Papers, 30, 198-204.
218. Kolarik, J. and Lednicky, F. (1986) Structure of polypropylene/EPDM
elastomer/calcium carbonate composites, in Polymer Composites (ed.
B. Sedlacek), Walter de Gruyter, Berlin, pp. 537-44.
219. Stamhuis, J. E. (1984) Polymer Composites, 5, 202-7.
220. Stamhuis, J. E. (1988) Polymer Composites, 9, 72-7.
221. Stamhuis, J. E. (1988) Polymer Composites, 9, 280-4.
222. Kitamura, H. (1982) Polypropylene composite for instrument panel of motor
vehicle, in Proceedings of the 4th International Conference on Composite
Materials, ICCM-IV, 25-27 October, Tokyo, Japan.
223. Kolarik, J., Lednicky, F. and Pukanszky, B. (1987) Ternary composites poly-
propylene/elastomer/filler: structure and elastic properties, in Proceedings of the
6th ICCM/2nd ECCM, Vol. 1 (eds F. L. Matthews, N. C. R. Buskell, J. M.
Hodgkinson and J. Morton), Elsevier, London, pp. 452-6l.
224. Pukcinszky, B., Tiid6s, F., Kolarik, J. and Lednicky, F. (1989) Composite
Polymers, 2, 491-511.
225. Kolarik, J., Pukanszky, B. and Lednicky F. (1988) Ternary composites poly-
propylene/elastomer/filler: Phase structure and its control, in Interfaces in
Polymer, Ceramic, and Metal Matrix Composites (ed. H. Ishida), Elsevier, New
York, pp. 453-62.
226. Kolarik, J., Lednicky, F. and Pukanszky, B. (1990) Composite Polymers, 3,
271-89.
227. Chiang, W.-Y., Yang, W.-D. , and Pukanszky, B. (1992) Polymer Engineering
and Science, 32, 641-8.
228. Kolarik, J., Lednicky, F., Jancar, J. and Pukanszky, B. (1990) Polymer Com-
munications, 31, 201-4.
229. Scott, C. and Ishida, H. (1987) Journal of Materials Science, 22, 3963-73.

You might also like